Unannounced changes to AI APIs
Price moves, context-window cuts, removed capabilities and vanished models — grouped so one upstream change is one row, and ranked by how big the change is rather than when we spotted it.
| Impact | Provider | What changed | Detected |
|---|---|---|---|
| truncation | gemini | -99% ctx · gemini/gemini-2.5-pro-preview-tts: context_tokens cut 1,048,576 to 8,192 | 2026-09-16 |
| truncation | bedrock | -98% max out · ai21.jamba-1-5-large-v1:0: max_output_tokens cut 256,000 to 4,096 +1 alias inferred | 2026-10-12 |
| truncation | openrouter | -98% max out · openrouter/moonshotai/kimi-k3:batch: max_output_tokens cut 943,718 to 16,384 | 2026-09-22 |
| cost | xiaomi | 34.29x price · xiaomi/mimo-v2.5-pro: cache_read_price_per_mtok $0.0028/Mtok to $0.10/Mtok | 2026-10-27 |
| truncation | bedrock_converse | -97% max out · zai.glm-4.7-flash: max_output_tokens cut 128,000 to 4,000 +1 alias | 2026-09-21 |
| truncation | thedrummer | -97% max out · thedrummer/unslopnemo-12b: max_output_tokens cut 1,024,000 to 32,768 inferred | 2026-08-24 |
| truncation | openai | -97% max out · openai/gpt-4.1-nano: max_output_tokens cut 942,818 to 32,768 | 2026-08-30 |
| truncation | nvidia | -96% max out · nvidia/nemotron-3-ultra-550b-a55b: max_output_tokens cut 461,059 to 16,384 | 2026-08-28 |
| truncation | azure_ai | -96% max out · azure_ai/grok-4.3: max_output_tokens cut 200,000 to 8,192 | 2026-09-28 |
| cost | qwen | 23.08x price · qwen/qwen3.8-27b: cache_read_price_per_mtok $0.0039/Mtok to $0.09/Mtok | 2026-10-28 |
| truncation | bedrock_converse | -95% max out · deepseek.v3.2: max_output_tokens cut 163,840 to 8,000 inferred | 2026-09-21 |
| truncation | bedrock | -95% max out · bedrock/ap-northeast-1/deepseek.v3.2: max_output_tokens cut 163,840 to 8,000 +7 aliases inferred | 2026-10-08 |
| cost | deepseek | 16.83x price · deepseek/deepseek-v4-flash-0731: cache_read_price_per_mtok $0.0014/Mtok to $0.02/Mtok | 2026-10-27 |
| truncation | bedrock_converse | -94% max out · moonshotai.kimi-k2.5: max_output_tokens cut 262,144 to 16,000 | 2026-09-21 |
| truncation | bedrock | -94% max out · bedrock/ap-northeast-1/moonshotai.kimi-k2-thinking: max_output_tokens cut 262,144 to 16,000 +15 aliases inferred | 2026-10-08 |
| truncation | qwen | -94% max out · qwen/qwen3-next-80b-a3b-instruct: max_output_tokens cut 262,144 to 16,384 +1 alias | 2026-08-19 |
| truncation | azure_ai | -94% max out · azure_ai/grok-4: max_output_tokens cut 131,072 to 8,192 +1 alias inferred | 2026-09-28 |
| truncation | bedrock_converse | -94% max out · qwen.qwen3-235b-a22b-2507-v1:0: max_output_tokens cut 131,072 to 8,192 | 2026-10-07 |
| truncation | watsonx | -94% max out · watsonx/meta-llama/llama-4-maverick-17b: max_output_tokens cut 128,000 to 8,192 inferred | 2026-09-06 |
| truncation | openrouter | -94% max out · openrouter/anthropic/claude-sonnet-4.5: max_output_tokens cut 1,000,000 to 64,000 | 2026-09-06 |
| truncation | vertex_ai | -94% ctx · vertex_ai/xai/grok-4.1-fast-non-reasoning: context_tokens cut 2,000,000 to 128,000 +1 alias inferred | 2026-09-10 |
| truncation | vertex_ai | -94% max out · vertex_ai/xai/grok-4.1-fast-non-reasoning: max_output_tokens cut 2,000,000 to 128,000 +1 alias inferred | 2026-09-10 |
| truncation | xiaomi | -93% max out · xiaomi/mimo-v2.5: max_output_tokens cut 945,000 to 65,536 inferred | 2026-10-27 |
| truncation | thinkingmachines | -93% max out · thinkingmachines/inkling: max_output_tokens cut 471,859 to 32,768 | 2026-09-10 |
| truncation | nvidia | -93% max out · nvidia/nemotron-3-super-120b-a12b: max_output_tokens cut 235,929 to 16,384 | 2026-10-08 |
| truncation | openrouter | -93% max out · openrouter/qwen/qwen3-235b-a22b-2507: max_output_tokens cut 235,929 to 16,384 +2 aliases | 2026-10-09 |
| truncation | qwen | -93% max out · qwen/qwen3-235b-a22b-2507: max_output_tokens cut 235,929 to 16,384 +8 aliases | 2026-10-21 |
| cost | qwen | 12.82x price · qwen/qwen3.8-27b: cache_read_price_per_mtok $0.0039/Mtok to $0.05/Mtok | 2026-10-29 |
| truncation | nvidia | -91% max out · nvidia/nemotron-3-ultra-550b-a55b: max_output_tokens cut 182,520 to 16,384 | 2026-10-02 |
| truncation | openrouter | -91% max out · openrouter/nvidia/nemotron-3-ultra-550b-a55b: max_output_tokens cut 182,520 to 16,384 | 2026-10-03 |
| truncation | novita | -91% ctx · novita/meta-llama/llama-3.3-70b-instruct: context_tokens cut 131,072 to 12,288 | 2026-08-28 |
| truncation | bedrock_converse | -90% max out · deepseek.v3-v1:0: max_output_tokens cut 81,920 to 8,000 inferred | 2026-10-08 |
| capability | arcee-ai | lost frequency_penalty, logit_bias, logprobs, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, top_logprobs · arcee-ai/trinity-large-thinking: capabilities lost frequency_penalty, logit_bias, logprobs, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, top_logprobs inferred | 2026-08-31 |
| capability | minimax | lost frequency_penalty, logprobs, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, top_k, top_logprobs · minimax/minimax-m2: capabilities lost frequency_penalty, logprobs, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, top_k, top_logprobs | 2026-10-28 |
| truncation | novita | -90% max out · novita/meta-llama/llama-3.3-70b-instruct: max_output_tokens cut 120,000 to 12,288 | 2026-08-28 |
| truncation | ~z-ai | -89% max out · ~z-ai/glm-flash-latest: max_output_tokens cut 943,718 to 102,400 | 2026-09-21 |
| cost | xiaomi | 8.40x price · xiaomi/mimo-v2.5: output_price_per_mtok $0.24/Mtok to $2.00/Mtok | 2026-10-27 |
| truncation | together_ai | -88% max out · together_ai/zai-org/GLM-5.2: max_output_tokens cut 1,048,575 to 128,000 +1 alias | 2026-08-30 |
| truncation | mistralai | -88% max out · mistralai/mistral-medium-3-5:batch: max_output_tokens cut 209,715 to 26,214 inferred | 2026-09-04 |
| truncation | fireworks_ai | -88% max out · fireworks_ai/accounts/fireworks/models/kimi-k2p5: max_output_tokens cut 262,144 to 32,768 +9 aliases inferred | 2026-07-31 |
| truncation | qwen | -88% max out · qwen/qwen3-coder-30b-a3b-instruct: max_output_tokens cut 262,144 to 32,768 +2 aliases | 2026-08-13 |
| capability | lost frequency_penalty, logprobs, presence_penalty, repetition_penalty, stop, structured_outputs, top_k, top_logprobs · google/gemma-4-26b-a4b-it:free: capabilities lost frequency_penalty, logprobs, presence_penalty, repetition_penalty, stop, structured_outputs, top_k, top_logprobs inferred | 2026-08-20 | |
| truncation | mistralai | -88% ctx · mistralai/mistral-medium-3-5:batch: context_tokens cut 262,144 to 32,768 inferred | 2026-09-04 |
| truncation | vertex_ai-language-models | -88% ctx · gemini-live-2.5-flash-native-audio: context_tokens cut 1,048,576 to 131,072 inferred | 2026-09-10 |
| truncation | z-ai | -88% max out · z-ai/glm-4.6: max_output_tokens cut 131,072 to 16,384 +3 aliases | 2026-09-19 |
| truncation | aion-labs | -88% ctx · aion-labs/aion-2.0: context_tokens cut 1,048,576 to 131,072 +2 aliases | 2026-09-22 |
| truncation | openrouter | -88% ctx · openrouter/aion-labs/aion-2.0: context_tokens cut 1,048,576 to 131,072 +2 aliases | 2026-09-23 |
| truncation | azure | -88% ctx · azure/eu/gpt-4o-realtime-preview-2024-12-17: context_tokens cut 128,000 to 16,000 +2 aliases | 2026-09-30 |
| capability | deepseek | lost logit_bias, logprobs, min_p, presence_penalty, repetition_penalty, seed, stop, top_logprobs · deepseek/deepseek-v3.2-exp: capabilities lost logit_bias, logprobs, min_p, presence_penalty, repetition_penalty, seed, stop, top_logprobs | 2026-09-30 |
| truncation | gemini | -88% ctx · gemini-2.5-flash-native-audio-latest: context_tokens cut 1,048,576 to 131,072 +5 aliases | 2026-10-01 |
| truncation | bedrock_converse | -88% max out · qwen.qwen3-coder-30b-a3b-v1:0: max_output_tokens cut 131,072 to 16,384 inferred | 2026-10-07 |
| truncation | openrouter | -87% max out · openrouter/z-ai/glm-4.6: max_output_tokens cut 131,000 to 16,384 | 2026-09-18 |
| truncation | openrouter | -87% max out · openrouter/z-ai/glm-5.2:free: max_output_tokens cut 230,400 to 29,491 | 2026-09-18 |
| truncation | meta-llama | -87% max out · meta-llama/llama-3.3-70b-instruct: max_output_tokens cut 128,000 to 16,384 | 2026-08-04 |
| truncation | openrouter | -87% ctx · openrouter/z-ai/glm-5.2:free: context_tokens cut 256,000 to 32,768 | 2026-09-18 |
| truncation | openrouter | -87% max out · openrouter/mistralai/mistral-small-3.2-24b-instruct: max_output_tokens cut 128,000 to 16,384 | 2026-09-18 |
| truncation | bedrock | -87% max out · bedrock/us-gov-east-1/openai.gpt-oss-120b-1:0: max_output_tokens cut 128,000 to 16,384 +3 aliases inferred | 2026-10-08 |
| truncation | bedrock_converse | -87% max out · openai.gpt-oss-120b-1:0: max_output_tokens cut 128,000 to 16,384 +3 aliases inferred | 2026-10-08 |
| truncation | bedrock | -87% max out · cohere.rerank-v3-5:0: max_output_tokens cut 32,000 to 4,096 inferred | 2026-10-12 |
| truncation | bedrock | -87% ctx · cohere.rerank-v3-5:0: context_tokens cut 32,000 to 4,096 inferred | 2026-10-12 |
| truncation | bedrock_converse | -87% ctx · us.writer.palmyra-x5-v1:0: context_tokens cut 1,000,000 to 128,000 +1 alias inferred | 2026-10-12 |
| truncation | ~z-ai | -86% max out · ~z-ai/glm-latest: max_output_tokens cut 943,718 to 128,000 +1 alias inferred | 2026-09-11 |
| truncation | openrouter | -86% max out · openrouter/qwen/qwen3-30b-a3b-instruct-2507: max_output_tokens cut 235,929 to 32,000 +1 alias | 2026-10-04 |
| truncation | qwen | -86% max out · qwen/qwen3-30b-a3b-instruct-2507: max_output_tokens cut 235,929 to 32,000 +6 aliases | 2026-10-07 |
| truncation | z-ai | -86% max out · z-ai/glm-5.2: max_output_tokens cut 943,718 to 131,072 +6 aliases | 2026-10-04 |
| truncation | openrouter | -86% max out · openrouter/z-ai/glm-5.2: max_output_tokens cut 943,718 to 131,072 +5 aliases | 2026-10-05 |
| truncation | ~deepseek | -86% max out · ~deepseek/deepseek-flash-latest: max_output_tokens cut 943,718 to 131,072 +4 aliases inferred | 2026-10-21 |
| truncation | reflection | -86% max out · reflection/beam-1.0-501b-a23b: max_output_tokens cut 943,718 to 131,072 | 2026-10-22 |
| truncation | ~z-ai | -86% max out · ~z-ai/glm-flash-latest: max_output_tokens cut 943,718 to 131,072 +20 aliases | 2026-10-24 |
| truncation | deepseek | -86% max out · deepseek/deepseek-v4-flash-0731: max_output_tokens cut 943,718 to 131,072 +4 aliases | 2026-10-27 |
| truncation | meta | -86% max out · meta/muse-spark-open: max_output_tokens cut 943,718 to 131,072 inferred | 2026-10-28 |
| truncation | openrouter | -86% max out · openrouter/z-ai/glm-5.3: max_output_tokens cut 943,717 to 131,072 | 2026-09-19 |
| truncation | deepseek | -86% max out · deepseek/deepseek-v4-flash: max_output_tokens cut 943,717 to 131,072 | 2026-10-27 |
| truncation | z-ai | -86% max out · z-ai/glm-5.3-flash: max_output_tokens cut 943,717 to 131,072 +8 aliases | 2026-10-30 |
| truncation | inclusionai | -86% max out · inclusionai/ling-3.0-flash-fin: max_output_tokens cut 235,929 to 32,768 | 2026-10-02 |
| truncation | openrouter | -86% max out · openrouter/inclusionai/ling-3.0-flash-fin: max_output_tokens cut 235,929 to 32,768 +3 aliases | 2026-10-03 |
| truncation | nvidia | -86% max out · nvidia/nemotron-3-nano-30b-a3b: max_output_tokens cut 235,929 to 32,768 | 2026-10-07 |
| truncation | qwen | -86% max out · qwen/qwen3-next-80b-a3b-instruct: max_output_tokens cut 235,929 to 32,768 +4 aliases | 2026-10-08 |
| truncation | -86% max out · google/gemma-4-26b-a4b-it: max_output_tokens cut 235,929 to 32,768 | 2026-10-12 | |
| truncation | meta | -86% max out · meta/muse-glimmer-30b: max_output_tokens cut 117,964 to 16,384 +2 aliases | 2026-10-06 |
| truncation | openrouter | -86% max out · openrouter/meta/muse-glimmer-30b: max_output_tokens cut 117,964 to 16,384 +1 alias | 2026-10-07 |
| truncation | qwen | -86% max out · qwen/qwen3-235b-a22b-thinking-2507: max_output_tokens cut 117,964 to 16,384 | 2026-10-09 |
| truncation | -86% max out · google/gemma-3-4b-it: max_output_tokens cut 117,964 to 16,384 | 2026-10-27 | |
| truncation | meta-llama | -86% max out · meta-llama/llama-3.1-8b-instruct: max_output_tokens cut 117,964 to 16,384 +1 alias | 2026-10-28 |
| cost | z-ai | 7.13x price · z-ai/glm-5.3-flash: output_price_per_mtok $0.07/Mtok to $0.50/Mtok | 2026-10-30 |
| truncation | meta-llama | -86% max out · meta-llama/llama-3.3-70b-instruct: max_output_tokens cut 115,200 to 16,384 +1 alias | 2026-09-14 |
| truncation | openrouter | -86% max out · openrouter/meta-llama/llama-4-maverick: max_output_tokens cut 115,200 to 16,384 | 2026-09-18 |
| truncation | qwen | -86% max out · qwen/qwen3.8-2.4t-a95b: max_output_tokens cut 909,000 to 131,072 | 2026-10-30 |
| cost | z-ai | 6.78x price · z-ai/glm-5.3-flash: output_price_per_mtok $0.07/Mtok to $0.50/Mtok | 2026-10-28 |
| cost | z-ai | 6.74x price · z-ai/glm-5.3-flash: output_price_per_mtok $0.07/Mtok to $0.50/Mtok | 2026-10-29 |
| truncation | meta-llama | -84% ctx · meta-llama/llama-guard-4-12b: context_tokens cut 1,048,576 to 163,840 | 2026-08-26 |
| capability | tencent | lost frequency_penalty, presence_penalty, repetition_penalty, response_format, stop, structured_outputs · tencent/hy3-preview: capabilities lost frequency_penalty, presence_penalty, repetition_penalty, response_format, stop, structured_outputs inferred | 2026-08-03 |
| capability | minimax | lost frequency_penalty, presence_penalty, repetition_penalty, seed, stop, top_k · minimax/minimax-m2.1: capabilities lost frequency_penalty, presence_penalty, repetition_penalty, seed, stop, top_k | 2026-10-08 |
| capability | ~deepseek | lost logit_bias, min_p, repetition_penalty, seed, structured_outputs, top_k · ~deepseek/deepseek-flash-latest: capabilities lost logit_bias, min_p, repetition_penalty, seed, structured_outputs, top_k inferred | 2026-10-17 |
| truncation | deepseek | -83% max out · deepseek/deepseek-v4-flash-0731: max_output_tokens cut 384,000 to 65,536 inferred | 2026-08-01 |
| truncation | nvidia | -82% max out · nvidia/nemotron-3-ultra-550b-a55b: max_output_tokens cut 182,520 to 32,768 +2 aliases | 2026-09-17 |
| cost | nvidia | 5.33x price · nvidia/nemotron-3-super-120b-a12b: output_price_per_mtok $0.45/Mtok to $2.40/Mtok +1 alias | 2026-10-28 |
| cost | qwen | 5.13x price · qwen/qwen3.8-27b: input_price_per_mtok $0.04/Mtok to $0.20/Mtok +1 alias | 2026-10-29 |
| truncation | bedrock_converse | -80% ctx · minimax.minimax-m2.5: context_tokens cut 1,000,000 to 196,000 inferred | 2026-09-22 |
| truncation | bedrock | -80% ctx · bedrock/ap-northeast-1/minimax.minimax-m2.5: context_tokens cut 1,000,000 to 196,000 +12 aliases inferred | 2026-10-08 |
| truncation | deepseek | -80% max out · deepseek/deepseek-v3.1-terminus: max_output_tokens cut 163,840 to 32,768 | 2026-08-20 |
| truncation | openrouter | -80% max out · openrouter/deepseek/deepseek-chat-v3.1: max_output_tokens cut 163,840 to 32,768 | 2026-09-18 |
| truncation | openrouter | -80% ctx · openrouter/anthropic/claude-sonnet-4.5: context_tokens cut 1,000,000 to 200,000 +1 alias | 2026-09-21 |
| capability | deepseek | lost logit_bias, min_p, presence_penalty, repetition_penalty, stop · deepseek/deepseek-chat-v3-0324: capabilities lost logit_bias, min_p, presence_penalty, repetition_penalty, stop | 2026-09-29 |
| capability | deepseek | lost logit_bias, min_p, presence_penalty, repetition_penalty, seed · deepseek/deepseek-v3.1-terminus: capabilities lost logit_bias, min_p, presence_penalty, repetition_penalty, seed | 2026-10-06 |
| capability | qwen | lost logprobs, presence_penalty, seed, stop, top_logprobs · qwen/qwen3-vl-30b-a3b-thinking: capabilities lost logprobs, presence_penalty, seed, stop, top_logprobs | 2026-10-09 |
| truncation | anthropic | -80% ctx · anthropic/claude-sonnet-4: context_tokens cut 1,000,000 to 200,000 +3 aliases | 2026-10-10 |
| truncation | bedrock | -80% ctx · anthropic.claude-3-5-sonnet-20240620-v1:0: context_tokens cut 1,000,000 to 200,000 +1 alias | 2026-10-13 |
| truncation | bedrock_converse | -80% ctx · anthropic.claude-sonnet-4-20250514-v1:0: context_tokens cut 1,000,000 to 200,000 +4 aliases | 2026-10-18 |
| capability | xiaomi | lost frequency_penalty, presence_penalty, repetition_penalty, stop, top_k · xiaomi/mimo-v2.5-pro: capabilities lost frequency_penalty, presence_penalty, repetition_penalty, stop, top_k inferred | 2026-10-23 |
| truncation | deepseek | -77% max out · deepseek/deepseek-chat-v3.1: max_output_tokens cut 144,900 to 32,768 +1 alias | 2026-09-08 |
| cost | ~deepseek | 4.29x price · ~deepseek/deepseek-flash-latest: output_price_per_mtok $0.23/Mtok to $1.00/Mtok | 2026-10-27 |
| truncation | nebius | -76% ctx · nebius/Qwen/Qwen2.5-VL-72B-Instruct: context_tokens cut 131,072 to 32,000 | 2026-09-06 |
| truncation | nebius | -76% max out · nebius/Qwen/Qwen2.5-VL-72B-Instruct: max_output_tokens cut 131,072 to 32,000 | 2026-09-06 |
| truncation | bedrock_mantle | -76% max out · bedrock_mantle/openai.gpt-oss-safeguard-120b: max_output_tokens cut 65,536 to 16,000 +1 alias inferred | 2026-10-08 |
| truncation | xiaomi | -75% max out · xiaomi/mimo-v2.5-pro: max_output_tokens cut 945,000 to 235,929 inferred | 2026-10-27 |
| truncation | wandb | -75% ctx · wandb/zai-org/GLM-5.2: context_tokens cut 1,049,000 to 262,000 inferred | 2026-10-30 |
| truncation | z-ai | -75% max out · z-ai/glm-5.3: max_output_tokens cut 943,718 to 235,929 | 2026-09-27 |
| truncation | ~z-ai | -75% max out · ~z-ai/glm-latest: max_output_tokens cut 943,718 to 235,929 +5 aliases | 2026-09-27 |
| truncation | fireworks_ai | -75% max out · fireworks_ai/accounts/fireworks/models/gpt-oss-120b: max_output_tokens cut 131,072 to 32,768 +1 alias | 2026-06-18 |
| truncation | liquid | -75% max out · liquid/lfm-2.5-2.6b:free: max_output_tokens cut 32,768 to 8,192 inferred | 2026-08-14 |
| truncation | ~deepseek | -75% max out · ~deepseek/deepseek-v4-flash-latest: max_output_tokens cut 1,048,576 to 262,144 +1 alias inferred | 2026-08-23 |
| truncation | qwen | -75% max out · qwen/qwen3.5-122b-a10b: max_output_tokens cut 262,144 to 65,536 +8 aliases | 2026-08-23 |
| truncation | qwen | -75% max out · qwen/qwen2.5-vl-72b-instruct: max_output_tokens cut 115,200 to 28,800 | 2026-08-26 |
| truncation | openai | -75% ctx · gpt-realtime-mini: context_tokens cut 128,000 to 32,000 | 2026-09-03 |
| capability | vertex_ai-language-models | lost pdf_input, prompt_caching, response_schema, url_context · gemini-live-2.5-flash-native-audio: capabilities lost pdf_input, prompt_caching, response_schema, url_context inferred | 2026-09-10 |
| capability | gemini | lost function_calling, response_schema, vision, web_search · gemini/gemini-2.5-pro-preview-tts: capabilities lost function_calling, response_schema, vision, web_search | 2026-09-16 |
| capability | replicate | lost function_calling, parallel_function_calling, response_schema, tool_choice · replicate/google/gemini-2.5-flash: capabilities lost function_calling, parallel_function_calling, response_schema, tool_choice; gained reasoning, video_input | 2026-09-19 |
| capability | replicate | lost function_calling, parallel_function_calling, response_schema, tool_choice · replicate/google/gemini-3-pro: capabilities lost function_calling, parallel_function_calling, response_schema, tool_choice; gained audio_input, video_input inferred | 2026-09-19 |
| truncation | openrouter | -75% max out · openrouter/qwen/qwen3.5-35b-a3b: max_output_tokens cut 65,536 to 16,384 | 2026-09-20 |
| truncation | qwen | -75% max out · qwen/qwen3.5-35b-a3b: max_output_tokens cut 65,536 to 16,384 | 2026-09-20 |
| capability | openai | lost frequency_penalty, logit_bias, presence_penalty, stop · openai/gpt-5-image-mini: capabilities lost frequency_penalty, logit_bias, presence_penalty, stop +2 aliases | 2026-10-02 |
| truncation | nvidia | -75% max out · nvidia/nemotron-3.5-lightning: max_output_tokens cut 131,072 to 32,768 +2 aliases | 2026-10-05 |
| truncation | openrouter | -75% max out · openrouter/nvidia/nemotron-3.5-lightning: max_output_tokens cut 131,072 to 32,768 | 2026-10-07 |
| truncation | bedrock_converse | -75% ctx · qwen.qwen3-32b-v1:0: context_tokens cut 131,072 to 32,768 | 2026-10-07 |
| truncation | bedrock_converse | -75% max out · qwen.qwen3-coder-480b-a35b-v1:0: max_output_tokens cut 65,536 to 16,384 inferred | 2026-10-07 |
| capability | qwen | lost logprobs, presence_penalty, seed, top_logprobs · qwen/qwen3-coder-30b-a3b-instruct: capabilities lost logprobs, presence_penalty, seed, top_logprobs | 2026-10-09 |
| truncation | gemini | -75% max out · gemini/gemini-2.5-pro-preview-tts: max_output_tokens cut 65,535 to 16,384 | 2026-09-16 |
| truncation | qwen | -75% max out · qwen/qwen3.6-27b: max_output_tokens cut 262,140 to 65,536 | 2026-10-07 |
| truncation | openrouter | -75% max out · openrouter/qwen/qwen3-coder: max_output_tokens cut 262,100 to 65,536 | 2026-09-18 |
| truncation | novita | -74% max out · novita/qwen/qwen2.5-7b-instruct: max_output_tokens cut 32,000 to 8,192 inferred | 2026-08-28 |
| truncation | bedrock_converse | -74% ctx · mistral.voxtral-mini-3b-2507: context_tokens cut 128,000 to 32,768 +1 alias inferred | 2026-10-08 |
| truncation | minimax | -74% max out · minimax/minimax-m2.5: max_output_tokens cut 128,000 to 32,768 | 2026-10-15 |
| truncation | azure | -74% ctx · azure/gpt-5.4-mini-2026-03-17: context_tokens cut 1,050,000 to 272,000 +3 aliases | 2026-07-30 |
| truncation | openai | -74% ctx · gpt-5.4-mini-2026-03-17: context_tokens cut 1,050,000 to 272,000 +3 aliases | 2026-07-30 |
| truncation | openrouter | -74% ctx · openrouter/nvidia/nemotron-3-super-120b-a12b: context_tokens cut 1,000,000 to 262,144 | 2026-09-18 |
| truncation | nvidia | -74% ctx · nvidia/nemotron-3-super-120b-a12b: context_tokens cut 1,000,000 to 262,144 +2 aliases | 2026-09-29 |
| cost | nvidia | 3.75x price · nvidia/nemotron-3-super-120b-a12b: input_price_per_mtok $0.08/Mtok to $0.30/Mtok +1 alias | 2026-10-28 |
| truncation | deepseek | -72% max out · deepseek/deepseek-v4-flash-vision-exp: max_output_tokens cut 943,718 to 262,144 +1 alias | 2026-09-27 |
| truncation | openrouter | -72% max out · openrouter/deepseek/deepseek-v4-flash-vision-exp: max_output_tokens cut 943,718 to 262,144 +1 alias | 2026-09-29 |
| truncation | z-ai | -72% max out · z-ai/glm-5.3: max_output_tokens cut 943,718 to 262,144 +1 alias | 2026-10-22 |
| truncation | deepseek | -72% max out · deepseek/deepseek-v4-flash-vision-exp: max_output_tokens cut 943,717 to 262,144 | 2026-09-28 |
| truncation | openrouter | -72% max out · openrouter/qwen/qwen3.5-35b-a3b: max_output_tokens cut 235,929 to 65,536 | 2026-10-09 |
| truncation | qwen | -72% max out · qwen/qwen3.5-122b-a10b: max_output_tokens cut 235,929 to 65,536 +10 aliases | 2026-10-29 |
| truncation | openai | -72% max out · openai/gpt-oss-20b: max_output_tokens cut 117,964 to 32,768 | 2026-09-22 |
| truncation | openrouter | -72% max out · openrouter/openai/gpt-oss-20b: max_output_tokens cut 117,964 to 32,768 | 2026-09-22 |
| cost | xiaomi | 3.36x price · xiaomi/mimo-v2.5: input_price_per_mtok $0.12/Mtok to $0.40/Mtok | 2026-10-27 |
| cost | z-ai | 3.30x price · z-ai/glm-5.3-flash: input_price_per_mtok $0.05/Mtok to $0.15/Mtok | 2026-10-28 |
| cost | z-ai | 3.28x price · z-ai/glm-5.3-flash: input_price_per_mtok $0.05/Mtok to $0.15/Mtok | 2026-10-29 |
| truncation | qwen | -69% max out · qwen/qwen3.5-122b-a10b: max_output_tokens cut 262,144 to 81,920 +1 alias | 2026-08-24 |
| truncation | qwen | -69% ctx · qwen/qwen3-14b: context_tokens cut 131,072 to 40,960 +1 alias | 2026-10-09 |
| truncation | qwen | -69% max out · qwen/qwen3.6-27b: max_output_tokens cut 262,140 to 81,920 | 2026-09-26 |
| truncation | openrouter | -69% max out · openrouter/qwen/qwen3.6-27b: max_output_tokens cut 262,140 to 81,920 | 2026-09-27 |
| truncation | openrouter | -68% max out · openrouter/anthropic/claude-haiku-4.5: max_output_tokens cut 200,000 to 64,000 | 2026-09-06 |
| cost | z-ai | 3.02x price · z-ai/glm-5.3-flash: input_price_per_mtok $0.05/Mtok to $0.15/Mtok | 2026-10-30 |
| capability | gemini | lost code_execution, file_search, service_tier · gemini/gemini-3.1-flash-lite-preview: capabilities lost code_execution, file_search, service_tier +1 alias | 2026-06-27 |
| capability | vertex_ai-language-models | lost code_execution, file_search, service_tier · gemini-3.1-flash-lite-preview: capabilities lost code_execution, file_search, service_tier +2 aliases | 2026-06-27 |
| capability | anthropic | lost max_completion_tokens, response_format, structured_outputs · anthropic/claude-opus-4.1: capabilities lost max_completion_tokens, response_format, structured_outputs | 2026-08-06 |
| capability | x-ai | lost frequency_penalty, presence_penalty, stop · x-ai/grok-4.3: capabilities lost frequency_penalty, presence_penalty, stop +3 aliases | 2026-08-18 |
| capability | ~x-ai | lost frequency_penalty, presence_penalty, stop · ~x-ai/grok-latest: capabilities lost frequency_penalty, presence_penalty, stop inferred | 2026-08-18 |
| capability | nvidia | lost logprobs, structured_outputs, top_logprobs · nvidia/nemotron-3-super-120b-a12b: capabilities lost logprobs, structured_outputs, top_logprobs | 2026-09-09 |
| capability | qwen | lost logit_bias, min_p, repetition_penalty · qwen/qwen3.8-flash: capabilities lost logit_bias, min_p, repetition_penalty | 2026-09-15 |
| capability | thedrummer | lost response_format, tool_choice, tools · thedrummer/unslopnemo-12b: capabilities lost response_format, tool_choice, tools inferred | 2026-09-16 |
| capability | gemini | lost function_calling, response_schema, web_search · gemini/gemini-2.5-flash-image: capabilities lost function_calling, response_schema, web_search | 2026-09-16 |
| capability | openrouter | lost function_calling, response_schema, tool_choice · openrouter/z-ai/glm-5.2:free: capabilities lost function_calling, response_schema, tool_choice | 2026-09-18 |
| truncation | ~deepseek | -67% max out · ~deepseek/deepseek-flash-latest: max_output_tokens cut 393,216 to 131,072 +2 aliases | 2026-09-18 |
| capability | minimax | lost logprobs, structured_outputs, top_logprobs · minimax/minimax-m2.7: capabilities lost logprobs, structured_outputs, top_logprobs | 2026-10-02 |
| capability | qwen | lost logprobs, repetition_penalty, top_logprobs · qwen/qwen3-235b-a22b-thinking-2507: capabilities lost logprobs, repetition_penalty, top_logprobs | 2026-10-08 |
| capability | inclusionai | lost logit_bias, min_p, structured_outputs · inclusionai/ling-3.0-flash-fin: capabilities lost logit_bias, min_p, structured_outputs | 2026-10-09 |
| capability | z-ai | lost logit_bias, min_p, repetition_penalty · z-ai/glm-5.3-flash:batch: capabilities lost logit_bias, min_p, repetition_penalty inferred | 2026-10-21 |
| truncation | deepseek | -67% max out · deepseek/deepseek-v4-flash: max_output_tokens cut 393,216 to 131,072 +2 aliases | 2026-10-27 |
| capability | qwen | lost logprobs, top_a, top_logprobs · qwen/qwen3-32b: capabilities lost logprobs, top_a, top_logprobs +3 aliases | 2026-10-28 |
| cost | deepseek | 2.96x price · deepseek/deepseek-v4-flash: cache_read_price_per_mtok $0.0095/Mtok to $0.03/Mtok | 2026-10-28 |
| cost | deepseek | 2.96x price · deepseek/deepseek-v4-flash: input_price_per_mtok $0.05/Mtok to $0.14/Mtok | 2026-10-28 |
| cost | deepseek | 2.96x price · deepseek/deepseek-v4-flash: output_price_per_mtok $0.09/Mtok to $0.28/Mtok | 2026-10-28 |
| truncation | arcee-ai | -66% max out · arcee-ai/trinity-large-thinking: max_output_tokens cut 235,929 to 80,000 inferred | 2026-08-30 |
| truncation | ~deepseek | -66% max out · ~deepseek/deepseek-v4-flash-latest: max_output_tokens cut 384,000 to 131,072 +1 alias | 2026-09-18 |
| truncation | deepseek | -66% max out · deepseek/deepseek-v4-flash-0731: max_output_tokens cut 384,000 to 131,072 +2 aliases | 2026-10-01 |
| truncation | openrouter | -66% max out · openrouter/deepseek/deepseek-v4-flash: max_output_tokens cut 384,000 to 131,072 | 2026-10-02 |
| truncation | openrouter | -65% max out · openrouter/qwen/qwen3.5-397b-a17b: max_output_tokens cut 235,929 to 81,920 | 2026-10-07 |
| truncation | qwen | -65% max out · qwen/qwen3.5-122b-a10b: max_output_tokens cut 235,929 to 81,920 +4 aliases | 2026-10-20 |
| truncation | nebius | -63% max out · nebius/deepseek-ai/DeepSeek-V4.1-Flash: max_output_tokens cut 1,048,576 to 384,000 | 2026-09-29 |
| cost | deepseek | 2.68x price · deepseek/deepseek-v4-flash-0731: input_price_per_mtok $0.04/Mtok to $0.12/Mtok | 2026-10-27 |
| truncation | novita | -62% max out · novita/moonshotai/kimi-k2-0905: max_output_tokens cut 262,144 to 100,352 +1 alias | 2026-08-28 |
| cost | z-ai | 2.57x price · z-ai/glm-5.3: output_price_per_mtok $1.40/Mtok to $3.60/Mtok | 2026-10-30 |
| truncation | deepseek | -61% ctx · deepseek/deepseek-r1: context_tokens cut 163,840 to 64,000 | 2026-08-13 |
| truncation | deepseek | -60% max out · deepseek/deepseek-v3.2: max_output_tokens cut 163,840 to 65,536 +1 alias | 2026-08-09 |
| truncation | novita | -60% max out · novita/deepseek/deepseek-v3-0324: max_output_tokens cut 163,840 to 65,536 inferred | 2026-08-28 |
| truncation | openrouter | -60% max out · openrouter/deepseek/deepseek-v3.2-exp: max_output_tokens cut 163,840 to 65,536 +1 alias | 2026-09-18 |
| cost | meta-llama | 2.50x price · meta-llama/llama-3.1-8b-instruct: input_price_per_mtok $0.02/Mtok to $0.05/Mtok | 2026-10-28 |
| cost | wandb | 2.50x price · wandb/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B: cache_read_price_per_mtok $0.02/Mtok to $0.05/Mtok | 2026-10-28 |
| truncation | openrouter | -59% max out · openrouter/~deepseek/deepseek-flash-latest: max_output_tokens cut 943,718 to 384,000 | 2026-09-21 |
| truncation | deepseek | -59% max out · deepseek/deepseek-v4-pro-0813: max_output_tokens cut 943,718 to 384,000 +3 aliases | 2026-09-28 |
| truncation | ~deepseek | -59% max out · ~deepseek/deepseek-flash-latest: max_output_tokens cut 943,718 to 384,000 +3 aliases | 2026-09-28 |
| truncation | deepseek | -59% max out · deepseek/deepseek-v4-pro-0813: max_output_tokens cut 943,717 to 384,000 | 2026-08-28 |
| truncation | novita | -59% max out · novita/qwen/qwen3-4b-fp8: max_output_tokens cut 20,000 to 8,192 inferred | 2026-08-28 |
| truncation | openrouter | -58% max out · openrouter/deepseek/deepseek-v4-pro-0813: max_output_tokens cut 943,718 to 393,216 +2 aliases | 2026-10-07 |
| truncation | deepseek | -58% max out · deepseek/deepseek-v4-pro-0813: max_output_tokens cut 943,718 to 393,216 +7 aliases | 2026-10-26 |
| truncation | ~deepseek | -58% max out · ~deepseek/deepseek-flash-latest: max_output_tokens cut 943,718 to 393,216 +21 aliases | 2026-10-26 |
| truncation | moonshotai | -58% max out · moonshotai/kimi-k2-thinking: max_output_tokens cut 235,929 to 98,304 +2 aliases | 2026-10-09 |
| truncation | moonshotai | -57% max out · moonshotai/kimi-k2-thinking: max_output_tokens cut 235,929 to 100,352 +2 aliases | 2026-09-13 |
| truncation | qwen | -56% max out · qwen/qwen3-14b: max_output_tokens cut 36,864 to 16,384 | 2026-10-10 |
| truncation | deepseek | -56% max out · deepseek/deepseek-v3.1-terminus: max_output_tokens cut 147,456 to 65,536 +6 aliases | 2026-10-17 |
| cost | ~moonshotai | 2.24x price · ~moonshotai/kimi-latest: output_price_per_mtok $8.94/Mtok to $20.00/Mtok | 2026-10-27 |
| truncation | sao10k | -55% max out · sao10k/l3-lunaris-8b: max_output_tokens cut 16,384 to 7,372 inferred | 2026-08-25 |
| truncation | openrouter | -55% max out · openrouter/qwen/qwen3-235b-a22b-thinking-2507: max_output_tokens cut 262,144 to 117,964 | 2026-09-18 |
| cost | deepseek | 2.20x price · deepseek/deepseek-v4.1-flash: output_price_per_mtok $0.19/Mtok to $0.42/Mtok | 2026-10-27 |
| truncation | openai | -53% max out · gpt-5-pro-2025-10-06: max_output_tokens cut 272,000 to 128,000 +1 alias | 2026-06-26 |
| truncation | azure_ai | -53% ctx · azure_ai/gpt-chat-latest: context_tokens cut 272,000 to 128,000 | 2026-10-24 |
| truncation | azure | -53% ctx · azure/gpt-chat-latest: context_tokens cut 272,000 to 128,000 +1 alias | 2026-10-24 |
| cost | mistral | 2.06x price · mistral/mistral-large-4-0: cache_read_price_per_mtok $0.07/Mtok to $0.14/Mtok +1 alias | 2026-10-28 |
| truncation | z-ai | -51% max out · z-ai/glm-5.2: max_output_tokens cut 262,144 to 128,000 | 2026-08-06 |
| cost | ~deepseek | 2.02x price · ~deepseek/deepseek-pro-latest: input_price_per_mtok $0.13/Mtok to $0.25/Mtok | 2026-10-28 |
| capability | vertex_ai-language-models | lost code_execution, file_search · vertex_ai/gemini-3.1-flash-lite-preview: capabilities lost code_execution, file_search | 2026-06-27 |
| capability | openrouter | lost code_execution, file_search · openrouter/google/gemini-3.1-flash-lite-preview: capabilities lost code_execution, file_search +1 alias | 2026-06-27 |
| truncation | xai | -50% max out · xai/grok-4.20-0309-reasoning: max_output_tokens cut 2,000,000 to 1,000,000 +3 aliases inferred | 2026-08-12 |
| truncation | xai | -50% ctx · xai/grok-4.20-0309-reasoning: context_tokens cut 2,000,000 to 1,000,000 +3 aliases inferred | 2026-08-12 |
| truncation | nvidia | -50% max out · nvidia/nemotron-3.5-lightning: max_output_tokens cut 262,144 to 131,072 | 2026-08-16 |
| truncation | qwen | -50% max out · qwen/qwen3.6-27b: max_output_tokens cut 131,072 to 65,536 | 2026-08-19 |
| truncation | mistral | -50% max out · mistral/codestral-2508: max_output_tokens cut 256,000 to 128,000 | 2026-08-21 |
| truncation | mistral | -50% ctx · mistral/codestral-2508: context_tokens cut 256,000 to 128,000 | 2026-08-21 |
| capability | qwen | lost logit_bias, min_p · qwen/qwen3-235b-a22b-thinking-2507: capabilities lost logit_bias, min_p | 2026-08-24 |
| truncation | qwen | -50% ctx · qwen/qwen3-235b-a22b-thinking-2507: context_tokens cut 262,144 to 131,072 | 2026-08-24 |
| truncation | -50% ctx · google/gemma-3-27b-it: context_tokens cut 262,144 to 131,072 | 2026-08-28 | |
| capability | xai | lost function_calling, tool_choice · xai/grok-4.20-multi-agent-0309: capabilities lost function_calling, tool_choice inferred | 2026-08-30 |
| capability | z-ai | lost logprobs, top_logprobs · z-ai/glm-4.7: capabilities lost logprobs, top_logprobs | 2026-08-31 |
| capability | kwaipilot | lost logprobs, top_logprobs · kwaipilot/kat-coder-pro-v2.5: capabilities lost logprobs, top_logprobs +1 alias inferred | 2026-08-31 |
| truncation | gryphe | -50% max out · gryphe/mythomax-l2-13b: max_output_tokens cut 7,372 to 3,686 | 2026-09-01 |
| capability | xiaomi | lost logprobs, top_logprobs · xiaomi/mimo-v2.5: capabilities lost logprobs, top_logprobs | 2026-09-02 |
| truncation | z-ai | -50% max out · z-ai/glm-5.2: max_output_tokens cut 262,144 to 131,072 +7 aliases | 2026-09-02 |
| capability | meta-llama | lost logprobs, top_logprobs · meta-llama/llama-3.1-70b-instruct: capabilities lost logprobs, top_logprobs | 2026-09-04 |
| truncation | watsonx | -50% max out · watsonx/bigscience/mt0-xxl-13b: max_output_tokens cut 8,192 to 4,096 inferred | 2026-09-06 |
| truncation | watsonx | -50% ctx · watsonx/bigscience/mt0-xxl-13b: context_tokens cut 8,192 to 4,096 inferred | 2026-09-06 |
| truncation | qwen | -50% max out · qwen/qwen3.6-27b: max_output_tokens cut 262,144 to 131,072 +5 aliases | 2026-09-09 |
| capability | z-ai | lost logit_bias, min_p · z-ai/glm-5: capabilities lost logit_bias, min_p | 2026-09-10 |
| truncation | openrouter | -50% max out · openrouter/qwen/qwen3.8-2.4t-a95b: max_output_tokens cut 262,144 to 131,072 | 2026-09-18 |
| truncation | openrouter | -50% ctx · openrouter/qwen/qwen3-235b-a22b-thinking-2507: context_tokens cut 262,144 to 131,072 | 2026-09-18 |
| truncation | cohere | -50% ctx · embed-english-light-v3.0: context_tokens cut 1,024 to 512 +3 aliases inferred | 2026-09-18 |
| truncation | meta-llama | -50% max out · meta-llama/llama-3.1-70b-instruct: max_output_tokens cut 16,384 to 8,192 +3 aliases | 2026-09-21 |
| capability | ~x-ai | lost stop, top_k · ~x-ai/grok-latest: capabilities lost stop, top_k | 2026-09-21 |
| capability | z-ai | lost logprobs, top_logprobs · z-ai/glm-5.3-flash:batch: capabilities lost logprobs, top_logprobs; gained min_p, seed +1 alias | 2026-09-22 |
| capability | thinkingmachines | lost logprobs, top_logprobs · thinkingmachines/inkling-small: capabilities lost logprobs, top_logprobs +1 alias | 2026-09-22 |
| truncation | thinkingmachines | -50% ctx · thinkingmachines/inkling-small: context_tokens cut 1,048,576 to 524,288 +1 alias | 2026-09-26 |
| truncation | openrouter | -50% max out · openrouter/meta-llama/llama-3.1-70b-instruct: max_output_tokens cut 16,384 to 8,192 +1 alias | 2026-09-27 |
| truncation | openrouter | -50% max out · openrouter/qwen/qwen3-vl-30b-a3b-instruct: max_output_tokens cut 32,768 to 16,384 | 2026-09-27 |
| capability | meta | lost logprobs, top_logprobs · meta/muse-glimmer-30b: capabilities lost logprobs, top_logprobs | 2026-09-28 |
| capability | xiaomi | lost logit_bias, min_p · xiaomi/mimo-v2.5-pro: capabilities lost logit_bias, min_p +1 alias | 2026-09-29 |
| truncation | openrouter | -50% ctx · openrouter/thinkingmachines/inkling-small: context_tokens cut 1,048,576 to 524,288 +1 alias | 2026-09-29 |
| truncation | qwen | -50% max out · qwen/qwen3-14b: max_output_tokens cut 16,384 to 8,192 +5 aliases | 2026-09-30 |
| capability | stepfun | lost logit_bias, min_p · stepfun/step-3.7-flash: capabilities lost logit_bias, min_p | 2026-10-01 |
| capability | ~anthropic | lost top_k, top_p · ~anthropic/claude-haiku-latest: capabilities lost top_k, top_p; gained reasoning_effort, verbosity | 2026-10-07 |
| truncation | bedrock_converse | -50% max out · qwen.qwen3-32b-v1:0: max_output_tokens cut 16,384 to 8,192 | 2026-10-07 |
| truncation | bedrock_mantle | -50% max out · bedrock_mantle/openai.gpt-oss-120b: max_output_tokens cut 32,768 to 16,384 +5 aliases inferred | 2026-10-08 |
| capability | sao10k | lost tool_choice, tools · sao10k/l3.1-euryale-70b: capabilities lost tool_choice, tools | 2026-10-08 |
| truncation | qwen | -50% max out · qwen/qwen3.5-35b-a3b: max_output_tokens cut 65,536 to 32,768 | 2026-10-09 |
| capability | qwen | lost response_format, seed · qwen/qwen3-235b-a22b-thinking-2507: capabilities lost response_format, seed | 2026-10-09 |
| truncation | qwen | -50% max out · qwen/qwen3-235b-a22b-2507: max_output_tokens cut 32,768 to 16,384 +5 aliases | 2026-10-13 |
| capability | inclusionai | lost logprobs, top_logprobs · inclusionai/ling-3.0-flash-vl: capabilities lost logprobs, top_logprobs inferred | 2026-10-13 |
| truncation | inclusionai | -50% ctx · inclusionai/ling-3.0-flash-vl: context_tokens cut 262,144 to 131,072 inferred | 2026-10-13 |
| truncation | meta-llama | -50% max out · meta-llama/llama-3.3-70b-instruct: max_output_tokens cut 32,768 to 16,384 | 2026-10-16 |
| capability | tencent | lost logprobs, top_logprobs · tencent/hy4-preview: capabilities lost logprobs, top_logprobs inferred | 2026-10-16 |
| truncation | meta-llama | -50% ctx · meta-llama/llama-4-maverick: context_tokens cut 1,048,576 to 524,288 | 2026-10-19 |
| capability | recursive | lost structured_outputs, tool_choice · recursive/flint-v1.0:free: capabilities lost structured_outputs, tool_choice inferred | 2026-10-22 |
| cost | 2.00x price · google/gemma-3-4b-it: output_price_per_mtok $0.05/Mtok to $0.10/Mtok | 2026-10-27 | |
| capability | openai | lost max_tokens, structured_outputs · openai/gpt-4: capabilities lost max_tokens, structured_outputs | 2026-10-28 |
| capability | qwen | lost logprobs, top_logprobs · qwen/qwen3-32b: capabilities lost logprobs, top_logprobs +1 alias | 2026-10-28 |
| capability | openai | lost max_tokens, prediction · openai/gpt-4o-2024-05-13: capabilities lost max_tokens, prediction | 2026-10-28 |
| capability | moonshotai | lost logprobs, top_logprobs · moonshotai/kimi-k2-thinking: capabilities lost logprobs, top_logprobs | 2026-10-28 |
| cost | meta-llama | 2.00x price · meta-llama/llama-3.1-8b-instruct: output_price_per_mtok $0.04/Mtok to $0.08/Mtok | 2026-10-28 |
| cost | wandb | 2.00x price · wandb/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B: output_price_per_mtok $0.75/Mtok to $1.50/Mtok | 2026-10-28 |
| cost | mistral | 2.00x price · mistral/mistral-large-4-0: output_price_per_mtok $2.09/Mtok to $4.18/Mtok +1 alias | 2026-10-28 |
| cost | mistral | 2.00x price · mistral/mistral-large-4-0: input_price_per_mtok $0.68/Mtok to $1.36/Mtok +1 alias | 2026-10-28 |
| capability | openrouter | lost prediction, web_search_options · openrouter/auto-beta: capabilities lost prediction, web_search_options; gained max_completion_tokens, parallel_tool_calls, verbosity inferred | 2026-10-29 |
| cost | qwen | 2.00x price · qwen/qwen3.8-27b: output_price_per_mtok $1.25/Mtok to $2.50/Mtok +1 alias | 2026-10-29 |
| truncation | bedrock | -50% max out · bedrock/eu-west-3/mistral.mistral-7b-instruct-v0:2: max_output_tokens cut 8,191 to 4,096 +12 aliases inferred | 2026-10-12 |
| truncation | bedrock_converse | -50% ctx · qwen.qwen3-coder-480b-a35b-v1:0: context_tokens cut 262,000 to 131,072 inferred | 2026-10-07 |
| truncation | nvidia | -49% ctx · nvidia/nemotron-3-ultra-550b-a55b: context_tokens cut 512,288 to 262,144 | 2026-08-28 |
| truncation | liquid | -49% ctx · liquid/lfm-2.5-2.6b:free: context_tokens cut 128,000 to 65,536 inferred | 2026-08-21 |
| truncation | mistralai | -49% ctx · mistralai/mistral-small-3.2-24b-instruct: context_tokens cut 256,000 to 131,072 | 2026-08-22 |
| truncation | watsonx | -49% max out · watsonx/mistralai/mistral-small-3-1-24b-instruct-2503: max_output_tokens cut 32,000 to 16,384 inferred | 2026-09-06 |
| cost | ~deepseek | 1.93x price · ~deepseek/deepseek-flash-latest: input_price_per_mtok $0.02/Mtok to $0.04/Mtok | 2026-10-29 |
| cost | ~deepseek | 1.92x price · ~deepseek/deepseek-pro-latest: cache_read_price_per_mtok $0.13/Mtok to $0.25/Mtok | 2026-10-28 |
| cost | ~moonshotai | 1.89x price · ~moonshotai/kimi-latest: output_price_per_mtok $10.56/Mtok to $20.00/Mtok | 2026-10-27 |
| cost | ~deepseek | 1.86x price · ~deepseek/deepseek-pro-latest: input_price_per_mtok $0.12/Mtok to $0.22/Mtok | 2026-10-27 |
| cost | deepseek | 1.85x price · deepseek/deepseek-v4-pro: cache_read_price_per_mtok $0.07/Mtok to $0.14/Mtok | 2026-10-29 |
| truncation | ~z-ai | -46% max out · ~z-ai/glm-latest: max_output_tokens cut 235,929 to 128,000 inferred | 2026-09-10 |
| cost | ~deepseek | 1.83x price · ~deepseek/deepseek-pro-latest: cache_read_price_per_mtok $0.12/Mtok to $0.22/Mtok | 2026-10-27 |
| cost | deepseek | 1.83x price · deepseek/deepseek-v4-pro: input_price_per_mtok $0.88/Mtok to $1.60/Mtok | 2026-10-29 |
| cost | deepseek | 1.83x price · deepseek/deepseek-v4-pro: output_price_per_mtok $1.75/Mtok to $3.20/Mtok | 2026-10-29 |
| truncation | thinkingmachines | -44% max out · thinkingmachines/inkling: max_output_tokens cut 471,859 to 262,144 | 2026-10-02 |
| truncation | openrouter | -44% max out · openrouter/thinkingmachines/inkling: max_output_tokens cut 471,859 to 262,144 | 2026-10-03 |
| truncation | openrouter | -44% max out · openrouter/~z-ai/glm-latest: max_output_tokens cut 235,929 to 131,072 | 2026-09-19 |
| truncation | nvidia | -44% max out · nvidia/nemotron-3.5-lightning: max_output_tokens cut 235,929 to 131,072 +2 aliases | 2026-09-26 |
| truncation | ~z-ai | -44% max out · ~z-ai/glm-latest: max_output_tokens cut 235,929 to 131,072 +2 aliases | 2026-09-27 |
| truncation | qwen | -44% max out · qwen/qwen3.8-27b: max_output_tokens cut 235,929 to 131,072 | 2026-09-30 |
| truncation | openrouter | -44% max out · openrouter/openai/gpt-oss-120b: max_output_tokens cut 117,964 to 65,536 | 2026-09-19 |
| truncation | openai | -44% max out · openai/gpt-oss-120b: max_output_tokens cut 117,964 to 65,536 +2 aliases | 2026-10-15 |
| cost | deepseek | 1.79x price · deepseek/deepseek-v4-flash: cache_read_price_per_mtok $0.03/Mtok to $0.05/Mtok | 2026-10-27 |
| cost | deepseek | 1.79x price · deepseek/deepseek-v4-flash-0731: output_price_per_mtok $0.13/Mtok to $0.24/Mtok | 2026-10-27 |
| cost | ~deepseek | 1.76x price · ~deepseek/deepseek-flash-latest: cache_read_price_per_mtok $0.02/Mtok to $0.03/Mtok | 2026-10-29 |
| cost | ~z-ai | 1.68x price · ~z-ai/glm-flash-latest: output_price_per_mtok $0.07/Mtok to $0.12/Mtok | 2026-10-28 |
| cost | qwen | 1.67x price · qwen/qwen3.6-35b-a3b: cache_read_price_per_mtok $0.03/Mtok to $0.05/Mtok | 2026-10-27 |
| cost | together_ai | 1.67x price · together_ai/Qwen/Qwen3.7-Max: output_price_per_mtok $4.50/Mtok to $7.50/Mtok +1 alias | 2026-10-28 |
| cost | wandb | 1.67x price · wandb/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B: input_price_per_mtok $0.15/Mtok to $0.25/Mtok | 2026-10-28 |
| cost | together_ai | 1.67x price · together_ai/Qwen/Qwen3.7-Max: input_price_per_mtok $1.50/Mtok to $2.50/Mtok +1 alias | 2026-10-28 |
| cost | together_ai | 1.67x price · together_ai/Qwen/Qwen3.7-Max: cache_read_price_per_mtok $0.30/Mtok to $0.50/Mtok +1 alias | 2026-10-28 |
| truncation | meta-llama | -39% max out · meta-llama/llama-3.2-3b-instruct: max_output_tokens cut 117,964 to 72,000 +1 alias | 2026-10-18 |
| cost | deepseek | 1.61x price · deepseek/deepseek-v4-flash: output_price_per_mtok $0.28/Mtok to $0.45/Mtok | 2026-10-27 |
| cost | tencent | 1.60x price · tencent/hy3: cache_read_price_per_mtok $0.02/Mtok to $0.03/Mtok +64 aliases | 2026-10-30 |
| cost | tencent | 1.60x price · tencent/hy3: output_price_per_mtok $0.33/Mtok to $0.53/Mtok +64 aliases | 2026-10-30 |
| cost | tencent | 1.60x price · tencent/hy3: input_price_per_mtok $0.08/Mtok to $0.13/Mtok +64 aliases | 2026-10-30 |
| cost | moonshotai | 1.58x price · moonshotai/kimi-k3: cache_read_price_per_mtok $0.14/Mtok to $0.22/Mtok | 2026-10-28 |
| cost | xiaomi | 1.58x price · xiaomi/mimo-v2.5-pro: input_price_per_mtok $0.30/Mtok to $0.48/Mtok | 2026-10-27 |
| truncation | openrouter | -36% max out · openrouter/minimax/minimax-m2: max_output_tokens cut 204,800 to 131,072 | 2026-09-18 |
| truncation | fireworks_ai | -35% max out · fireworks_ai/accounts/fireworks/models/glm-5p1: max_output_tokens cut 202,800 to 131,072 +1 alias inferred | 2026-06-18 |
| cost | moonshotai | 1.54x price · moonshotai/kimi-k3: input_price_per_mtok $0.15/Mtok to $0.23/Mtok | 2026-10-28 |
| cost | deepseek | 1.54x price · deepseek/deepseek-v4-flash: input_price_per_mtok $0.09/Mtok to $0.14/Mtok | 2026-10-27 |
| cost | deepseek | 1.54x price · deepseek/deepseek-v4-flash: cache_read_price_per_mtok $0.02/Mtok to $0.03/Mtok | 2026-10-27 |
| truncation | undi95 | -33% max out · undi95/remm-slerp-l2-13b: max_output_tokens cut 6,144 to 4,096 | 2026-08-24 |
| cost | ~moonshotai | 1.50x price · ~moonshotai/kimi-latest: cache_read_price_per_mtok $0.14/Mtok to $0.21/Mtok | 2026-10-28 |
| cost | ~deepseek | 1.49x price · ~deepseek/deepseek-flash-latest: output_price_per_mtok $1.21/Mtok to $1.80/Mtok | 2026-10-27 |
| truncation | azure_ai | -32% ctx · azure_ai/gpt-5.4-mini-2026-03-17: context_tokens cut 400,000 to 272,000 +3 aliases | 2026-07-30 |
| truncation | ~deepseek | -32% max out · ~deepseek/deepseek-v4-flash-latest: max_output_tokens cut 384,000 to 262,144 +3 aliases inferred | 2026-08-19 |
| cost | ~z-ai | 1.46x price · ~z-ai/glm-flash-latest: input_price_per_mtok $0.03/Mtok to $0.04/Mtok | 2026-10-27 |
| cost | ~moonshotai | 1.46x price · ~moonshotai/kimi-latest: input_price_per_mtok $0.15/Mtok to $0.22/Mtok | 2026-10-28 |
| cost | z-ai | 1.45x price · z-ai/glm-5.1: output_price_per_mtok $3.04/Mtok to $4.40/Mtok +1 alias | 2026-10-28 |
| cost | z-ai | 1.45x price · z-ai/glm-5.1: input_price_per_mtok $0.97/Mtok to $1.40/Mtok +1 alias | 2026-10-28 |
| cost | z-ai | 1.45x price · z-ai/glm-5.1: cache_read_price_per_mtok $0.18/Mtok to $0.26/Mtok +1 alias | 2026-10-28 |
| cost | ~deepseek | 1.45x price · ~deepseek/deepseek-flash-latest: cache_read_price_per_mtok $0.01/Mtok to $0.02/Mtok | 2026-10-29 |
| cost | deepseek | 1.43x price · deepseek/deepseek-v4.1-flash: cache_read_price_per_mtok $0.0042/Mtok to $0.0060/Mtok +1 alias | 2026-10-28 |
| cost | ~deepseek | 1.43x price · ~deepseek/deepseek-v4-flash-latest: input_price_per_mtok $0.04/Mtok to $0.05/Mtok | 2026-10-30 |
| truncation | z-ai | -30% max out · z-ai/glm-5.1: max_output_tokens cut 182,476 to 128,000 | 2026-08-30 |
| cost | qwen | 1.41x price · qwen/qwen3.6-27b: input_price_per_mtok $0.32/Mtok to $0.45/Mtok | 2026-10-28 |
| truncation | z-ai | -28% max out · z-ai/glm-5.2: max_output_tokens cut 182,476 to 131,072 | 2026-09-14 |
| cost | moonshotai | 1.39x price · moonshotai/kimi-k3: output_price_per_mtok $10.80/Mtok to $15.00/Mtok | 2026-10-30 |
| truncation | minimax | -26% max out · minimax/minimax-m2.7: max_output_tokens cut 176,947 to 131,072 +4 aliases | 2026-10-14 |
| cost | deepseek | 1.35x price · deepseek/deepseek-v4-pro: cache_read_price_per_mtok $0.06/Mtok to $0.08/Mtok | 2026-10-27 |
| cost | deepseek | 1.35x price · deepseek/deepseek-v4-pro: output_price_per_mtok $1.42/Mtok to $1.91/Mtok | 2026-10-27 |
| cost | ~z-ai | 1.33x price · ~z-ai/glm-flash-latest: input_price_per_mtok $0.03/Mtok to $0.03/Mtok | 2026-10-28 |
| cost | ~deepseek | 1.33x price · ~deepseek/deepseek-flash-latest: cache_read_price_per_mtok $0.01/Mtok to $0.02/Mtok | 2026-10-27 |
| cost | z-ai | 1.32x price · z-ai/glm-5.3-flash: input_price_per_mtok $0.03/Mtok to $0.05/Mtok | 2026-10-29 |
| cost | ~z-ai | 1.32x price · ~z-ai/glm-flash-latest: input_price_per_mtok $0.03/Mtok to $0.05/Mtok | 2026-10-29 |
| cost | ~z-ai | 1.32x price · ~z-ai/glm-flash-latest: output_price_per_mtok $0.07/Mtok to $0.09/Mtok +1 alias | 2026-10-30 |
| truncation | novita | -23% max out · novita/moonshotai/kimi-k2-instruct: max_output_tokens cut 131,072 to 100,352 inferred | 2026-08-28 |
| cost | z-ai | 1.30x price · z-ai/glm-5.2: output_price_per_mtok $5.40/Mtok to $7.00/Mtok | 2026-10-29 |
| truncation | openrouter | -22% max out · openrouter/mistralai/mistral-small-3.1-24b-instruct: max_output_tokens cut 131,072 to 102,400 +2 aliases | 2026-09-21 |
| truncation | z-ai | -22% max out · z-ai/glm-5.3-flash: max_output_tokens cut 131,072 to 102,400 | 2026-09-21 |
| truncation | azure_ai | -22% max out · azure_ai/deepseek-v3.2-speciale: max_output_tokens cut 163,840 to 128,000 +1 alias | 2026-09-28 |
| truncation | azure_ai | -22% ctx · azure_ai/deepseek-v3.2-speciale: context_tokens cut 163,840 to 128,000 +1 alias | 2026-09-29 |
| truncation | deepseek | -22% max out · deepseek/deepseek-chat-v3-0324: max_output_tokens cut 147,456 to 115,200 +1 alias | 2026-10-05 |
| truncation | openrouter | -22% max out · openrouter/deepseek/deepseek-chat-v3-0324: max_output_tokens cut 147,456 to 115,200 | 2026-10-07 |
| truncation | bedrock_converse | -22% ctx · deepseek.v3-v1:0: context_tokens cut 163,840 to 128,000 inferred | 2026-10-08 |
| truncation | thedrummer | -20% max out · thedrummer/unslopnemo-12b: max_output_tokens cut 32,768 to 26,214 inferred | 2026-08-25 |
| truncation | openrouter | -20% max out · openrouter/mistralai/mixtral-8x22b-instruct: max_output_tokens cut 65,536 to 52,428 | 2026-09-18 |
| truncation | openrouter | -20% max out · openrouter/mistralai/ministral-3b-2512: max_output_tokens cut 131,072 to 104,857 | 2026-09-18 |
| truncation | openrouter | -20% max out · openrouter/mistralai/ministral-14b-2512: max_output_tokens cut 262,144 to 209,715 +2 aliases | 2026-09-18 |
| truncation | mistralai | -20% max out · mistralai/mistral-small-3.1-24b-instruct: max_output_tokens cut 128,000 to 102,400 | 2026-08-25 |
| truncation | ~z-ai | -20% ctx · ~z-ai/glm-latest: context_tokens cut 1,310,720 to 1,048,576 inferred | 2026-09-09 |
| truncation | z-ai | -20% ctx · z-ai/glm-5.3: context_tokens cut 1,310,720 to 1,048,576 | 2026-09-09 |
| truncation | z-ai | -20% max out · z-ai/glm-5.2: max_output_tokens cut 163,840 to 131,072 | 2026-09-18 |
| truncation | openrouter | -20% ctx · openrouter/deepseek/deepseek-v4-flash-0731: context_tokens cut 1,310,720 to 1,048,576 +1 alias | 2026-10-02 |
| truncation | openrouter | -20% max out · openrouter/qwen/qwen3.6-27b: max_output_tokens cut 81,920 to 65,536 | 2026-10-07 |
| truncation | qwen | -20% max out · qwen/qwen3.5-122b-a10b: max_output_tokens cut 81,920 to 65,536 +3 aliases | 2026-10-28 |
| cost | ~deepseek | 1.25x price · ~deepseek/deepseek-flash-latest: output_price_per_mtok $1.44/Mtok to $1.80/Mtok | 2026-10-29 |
| truncation | nebius | -14% max out · nebius/google/gemma-3-27b-it: max_output_tokens cut 128,000 to 110,000 | 2026-09-06 |
| truncation | nebius | -14% ctx · nebius/google/gemma-3-27b-it: context_tokens cut 128,000 to 110,000 | 2026-09-06 |
| truncation | nvidia | -13% max out · nvidia/nemotron-3-nano-30b-a3b: max_output_tokens cut 262,144 to 228,000 inferred | 2026-08-10 |
| truncation | openrouter | -13% max out · openrouter/qwen/qwen-2.5-coder-32b-instruct: max_output_tokens cut 33,792 to 29,491 | 2026-09-18 |
| truncation | openai | -12% ctx · daybreak-blue-latest: context_tokens cut 1,050,000 to 922,000 +5 aliases | 2026-09-17 |
| truncation | azure | -12% ctx · azure/eu/gpt-5.4-2026-03-05: context_tokens cut 1,050,000 to 922,000 +28 aliases | 2026-09-30 |
| truncation | azure_ai | -12% ctx · azure_ai/gpt-5.4-2026-03-05: context_tokens cut 1,050,000 to 922,000 +6 aliases | 2026-09-30 |
| truncation | bedrock_mantle | -12% ctx · bedrock_mantle/openai.gpt-5.4: context_tokens cut 1,050,000 to 922,000 +11 aliases inferred | 2026-10-27 |
| truncation | bedrock_converse | -12% ctx · global.openai.gpt-5.4: context_tokens cut 1,050,000 to 922,000 +17 aliases | 2026-10-27 |
| truncation | openai | -10% max out · openai/gpt-3.5-turbo-0613: max_output_tokens cut 4,096 to 3,685 +1 alias | 2026-08-25 |
| truncation | gryphe | -10% max out · gryphe/mythomax-l2-13b: max_output_tokens cut 4,096 to 3,686 | 2026-08-25 |
| truncation | deepseek | -10% max out · deepseek/deepseek-r1-distill-llama-70b: max_output_tokens cut 8,192 to 7,372 | 2026-08-25 |
| truncation | microsoft | -10% max out · microsoft/phi-4: max_output_tokens cut 16,384 to 14,745 | 2026-08-25 |
| truncation | rekaai | -10% max out · rekaai/reka-edge: max_output_tokens cut 16,384 to 14,745 inferred | 2026-08-25 |
| truncation | z-ai | -10% max out · z-ai/glm-5.1: max_output_tokens cut 202,752 to 182,476 | 2026-08-25 |
| truncation | openrouter | -10% max out · openrouter/moonshotai/kimi-k2.5: max_output_tokens cut 262,144 to 235,929 +1 alias | 2026-09-18 |
| truncation | qwen | -10% max out · qwen/qwen3.8-27b: max_output_tokens cut 262,144 to 235,929 | 2026-10-30 |
| truncation | minimax | -10% max out · minimax/minimax-01: max_output_tokens cut 1,000,192 to 900,172 inferred | 2026-08-25 |
| truncation | stepfun | -10% max out · stepfun/step-3.7-flash: max_output_tokens cut 256,000 to 230,400 inferred | 2026-08-25 |
| truncation | dots-studio | -10% max out · dots-studio/dots-3-note-preview:free: max_output_tokens cut 512,000 to 460,800 inferred | 2026-08-25 |
| truncation | z-ai | -10% max out · z-ai/glm-5.2:free: max_output_tokens cut 256,000 to 230,400 inferred | 2026-08-25 |
| truncation | ibm-granite | -10% max out · ibm-granite/granite-4.0-h-micro: max_output_tokens cut 131,000 to 117,900 | 2026-08-25 |
| truncation | deepseek | -10% max out · deepseek/deepseek-chat-v3-0324: max_output_tokens cut 163,840 to 147,456 +1 alias | 2026-08-25 |
| truncation | meta-llama | -10% max out · meta-llama/llama-3.2-1b-instruct: max_output_tokens cut 60,000 to 54,000 | 2026-08-25 |
| truncation | qwen | -10% max out · qwen/qwen2.5-vl-72b-instruct: max_output_tokens cut 128,000 to 115,200 | 2026-08-25 |
| truncation | deepseek | -10% max out · deepseek/deepseek-chat-v3.1: max_output_tokens cut 161,000 to 144,900 | 2026-08-25 |
Announced deprecations
Changes that came with a published date. Separated out because they are a different problem: you were told.
| Impact | Provider | What changed | Detected |
|---|---|---|---|
| availability | databricks | databricks/databricks-deepseek-v4-flash-0731: deprecation announced for 2026-11-05 — in 6 days inferred | 2026-10-30 |
| availability | databricks | databricks/databricks-deepseek-v4-pro-0813: deprecation announced for 2026-10-30 — in 0 days +1 alias inferred | 2026-10-30 |
| availability | databricks | databricks/databricks-deepseek-v4-flash-0731: retires 2026-11-05 (in 6 days) inferred | 2026-10-30 |
| availability | databricks | databricks/databricks-deepseek-v4-pro-0813: retires 2026-10-30 (in 0 days) +1 alias inferred | 2026-10-30 |
| availability | databricks | databricks/databricks-deepseek-v4-1-flash: retires 2026-11-19 (in 22 days) inferred | 2026-10-28 |
| availability | scaleway | scaleway/glm-5.2: retires 2026-11-16 (in 19 days) | 2026-10-28 |
| availability | scaleway | scaleway/glm-5.2: deprecation announced for 2026-11-16 — in 19 days | 2026-10-28 |
| availability | azure_ai | azure_ai/Llama-4-Maverick-17B-128E-Instruct-FP8: deprecation announced for 2026-11-19 — in 22 days inferred | 2026-10-28 |
| availability | azure_ai | azure_ai/Llama-4-Maverick-17B-128E-Instruct-FP8: retires 2026-11-19 (in 22 days) inferred | 2026-10-28 |
| availability | together_ai | together_ai/MiniMaxAI/MiniMax-M3: retires 2026-11-06 (in 10 days) +1 alias | 2026-10-27 |
| availability | together_ai | together_ai/MiniMaxAI/MiniMax-M3: deprecation announced for 2026-11-06 — in 10 days +1 alias | 2026-10-27 |
| availability | meta-llama | meta-llama/llama-3.1-70b-instruct: retires 2026-11-03 (in 7 days) | 2026-10-27 |
| availability | meta-llama | meta-llama/llama-3.1-70b-instruct: retirement announced for 2026-11-03 — in 7 days | 2026-10-27 |
| availability | upstage | upstage/solar-pro-3: retirement announced for 2026-10-30 — in 4 days inferred | 2026-10-26 |
| availability | upstage | upstage/solar-pro-3: retires 2026-10-30 (in 4 days) inferred | 2026-10-26 |
| availability | github_copilot | github_copilot/claude-opus-4.8-fast: deprecation announced for 2026-11-16 — in 23 days +3 aliases | 2026-10-24 |
| availability | github_copilot | github_copilot/claude-opus-4.8-fast: retires 2026-11-16 (in 23 days) +3 aliases | 2026-10-24 |
| availability | arcee-ai | arcee-ai/trinity-large-thinking: retirement announced for 2026-10-26 — in 4 days inferred | 2026-10-22 |
| availability | arcee-ai | arcee-ai/trinity-large-thinking: retires 2026-10-26 (in 4 days) inferred | 2026-10-22 |
| availability | fireworks_ai | fireworks_ai/accounts/fireworks/models/gpt-oss-120b: retires 2026-11-07 (in 18 days) +8 aliases | 2026-10-20 |
| availability | fireworks_ai | fireworks_ai/accounts/fireworks/models/gpt-oss-120b: deprecation announced for 2026-11-07 — in 18 days +8 aliases | 2026-10-20 |
| availability | veo-3.1-generate-preview: retires 2026-10-22 (in 2 days) +1 alias inferred | 2026-10-20 | |
| availability | github_copilot | github_copilot/gemini-3.7-flash: retires 2026-10-19 (1 days ago) +5 aliases | 2026-10-20 |
| availability | github_copilot | github_copilot/gemini-3.7-flash: deprecation announced for 2026-10-19 (date already passed) +5 aliases | 2026-10-20 |
| availability | baseten | baseten/MiniMaxAI/MiniMax-M2.5: retires 2026-10-16 (3 days ago) +14 aliases | 2026-10-19 |
| availability | nousresearch | nousresearch/hermes-3-llama-3.1-405b: retirement announced for 2026-11-02 — in 14 days +1 alias | 2026-10-19 |
| availability | baseten | baseten/MiniMaxAI/MiniMax-M2.5: deprecation announced for 2026-10-16 (date already passed) +14 aliases | 2026-10-19 |
| availability | nousresearch | nousresearch/hermes-3-llama-3.1-405b: retires 2026-11-02 (in 14 days) +1 alias | 2026-10-19 |
| availability | openai | gpt-4-1106-preview: retirement announced for 2026-10-23 — in 8 days inferred | 2026-10-15 |
| availability | anthropic | anthropic/claude-sonnet-4: retirement announced for 2026-10-15 — in 0 days | 2026-10-15 |
| availability | anthropic | anthropic/claude-sonnet-4: retires 2026-10-15 (in 0 days) +2 aliases | 2026-10-15 |
| availability | unbiased | unbiased/pareto-26.10-preview: retirement announced for 2026-10-15 — in 1 days inferred | 2026-10-14 |
| availability | unbiased | unbiased/pareto-26.10-preview: retires 2026-10-15 (in 1 days) inferred | 2026-10-14 |
| availability | openai | gpt-4o-mini-tts-2025-03-20: deprecation announced for 2026-10-01 (date already passed) +6 aliases inferred | 2026-10-13 |
| availability | together_ai | together_ai/deepseek-ai/DeepSeek-V4-Flash-0731: deprecation announced for 2026-10-22 — in 13 days +3 aliases | 2026-10-13 |
| availability | antigravity-preview-09-2026: retires 2026-09-30 (10 days ago) inferred | 2026-10-10 | |
| availability | antigravity-preview-09-2026: retirement announced for 2026-09-30 (date already passed) inferred | 2026-10-10 | |
| availability | xiaomi | xiaomi/mimo-v2.5-pro: retires 2026-10-21 (in 11 days) +1 alias inferred | 2026-10-10 |
| availability | xiaomi | xiaomi/mimo-v2.5-pro: retirement announced for 2026-10-21 — in 11 days +1 alias inferred | 2026-10-10 |
| availability | together_ai | together_ai/deepseek-ai/DeepSeek-V4-Flash-0731: retires 2026-10-22 (in 13 days) +1 alias | 2026-10-09 |
| availability | azure_ai | azure_ai/kimi-k2.7-code: deprecation announced for 2026-10-03 (date already passed) | 2026-10-09 |
| availability | gemini | gemini/deep-research-pro-preview-12-2025: retires 2026-10-23 (in 14 days) inferred | 2026-10-09 |
| availability | gemini | gemini/deep-research-pro-preview-12-2025: deprecation announced for 2026-10-23 — in 14 days inferred | 2026-10-09 |
| availability | openrouter | openrouter/qwen/qwen3.5-9b: retires 2026-10-21 (in 12 days) | 2026-10-09 |
| availability | openrouter | openrouter/qwen/qwen3.5-9b: deprecation announced for 2026-10-21 — in 12 days | 2026-10-09 |
| availability | gemini | gemini/gemini-omni-flash-preview: deprecation announced for 2026-10-22 — in 14 days | 2026-10-08 |
| availability | gemini | gemini/gemini-omni-flash-preview: retires 2026-10-22 (in 14 days) +3 aliases | 2026-10-08 |
| availability | azure | azure/sora-2: deprecation announced for 2026-11-02 — in 25 days inferred | 2026-10-08 |
| availability | azure | azure/sora-2: retires 2026-11-02 (in 25 days) inferred | 2026-10-08 |
| availability | qwen | qwen/qwen3-next-80b-a3b-instruct: retirement announced for 2026-10-08 — in 0 days | 2026-10-08 |
| availability | qwen | qwen/qwen3-next-80b-a3b-instruct: retires 2026-10-08 (in 0 days) | 2026-10-08 |
| availability | qwen | qwen/qwen3.5-9b: retirement announced for 2026-10-21 — in 14 days | 2026-10-07 |
| availability | qwen | qwen/qwen3.5-9b: retires 2026-10-21 (in 14 days) | 2026-10-07 |
| availability | gemini | gemini/gemini-3.1-flash-image: retires 2026-10-29 (in 23 days) | 2026-10-06 |
| availability | gemini | gemini/gemini-3.1-flash-image: deprecation announced for 2026-10-29 — in 23 days | 2026-10-06 |
| availability | azure | azure/gpt-4o-transcribe: deprecation announced for 2026-10-15 — in 10 days +1 alias | 2026-10-05 |
| availability | qwen | qwen/qwen-plus-2025-07-28: retirement announced for 2026-10-09 — in 14 days +16 aliases | 2026-10-03 |
| availability | openrouter | openrouter/qwen/qwen-plus-2025-07-28: retires 2026-10-09 (in 13 days) +16 aliases | 2026-10-03 |
| availability | openrouter | openrouter/poolside/laguna-s-2.1: retires 2026-10-31 (in 28 days) +3 aliases | 2026-10-03 |
| availability | openrouter | openrouter/poolside/laguna-s-2.1: deprecation announced for 2026-10-31 — in 28 days +3 aliases | 2026-10-03 |
| availability | openrouter | openrouter/qwen/qwen-plus-2025-07-28: deprecation announced for 2026-10-09 — in 13 days +16 aliases | 2026-10-03 |
| availability | qwen | qwen/qwen-plus-2025-07-28: retires 2026-10-09 (in 14 days) +15 aliases | 2026-10-03 |
| availability | xai | xai/grok-voice-transcribe-1.0: retires 2026-10-02 (in 0 days) inferred | 2026-10-02 |
| availability | xai | xai/grok-voice-transcribe-1.0: deprecation announced for 2026-10-02 — in 0 days inferred | 2026-10-02 |
| availability | poolside | poolside/laguna-s-2.1: retires 2026-10-31 (in 29 days) +3 aliases | 2026-10-02 |
| availability | poolside | poolside/laguna-s-2.1: retirement announced for 2026-10-31 — in 29 days +3 aliases | 2026-10-02 |
| availability | azure | azure/eu/gpt-4.1-nano: deprecation announced for 2026-10-14 — in 13 days +12 aliases | 2026-10-01 |
| availability | openrouter | openrouter/stealth/space-bunny-alpha: deprecation announced for 2026-10-05 — in 4 days | 2026-10-01 |
| availability | stealth | stealth/space-bunny-alpha: retires 2026-10-05 (in 4 days) | 2026-10-01 |
| availability | stealth | stealth/space-bunny-alpha: retirement announced for 2026-10-05 — in 4 days | 2026-10-01 |
| availability | openrouter | openrouter/stealth/space-bunny-alpha: retires 2026-10-05 (in 4 days) | 2026-10-01 |
| availability | mistral | mistral/labs-leanstral-1-5: deprecation announced for 2026-09-30 — in 0 days +1 alias inferred | 2026-09-30 |
| availability | mistral | mistral/labs-leanstral-1-5: retires 2026-09-30 (in 0 days) +1 alias inferred | 2026-09-30 |
| availability | gemini | gemini/veo-3.1-fast-generate-preview: deprecation announced for 2026-10-22 — in 22 days +2 aliases | 2026-09-30 |
| availability | together_ai | together_ai/google/gemma-4-31B-it: deprecation announced for 2026-09-15 (date already passed) +5 aliases | 2026-09-28 |
| availability | together_ai | together_ai/deepseek-ai/DeepSeek-V4-Pro: deprecation announced for 2026-08-27 (date already passed) +4 aliases | 2026-09-25 |
| availability | together_ai | together_ai/google/gemma-4-31B-it: deprecation announced for 2026-09-14 (date already passed) +5 aliases | 2026-09-25 |
| availability | antigravity-preview-09-2026: retires 2026-09-01 (23 days ago) inferred | 2026-09-24 | |
| availability | azure_ai | azure_ai/Cohere-command-a-plus-05-2026: deprecation announced for 2026-10-16 — in 22 days inferred | 2026-09-24 |
| availability | azure_ai | azure_ai/Cohere-command-a-plus-05-2026: retires 2026-10-16 (in 22 days) inferred | 2026-09-24 |
| availability | vertex_ai-moonshot_models | vertex_ai/moonshotai/kimi-k2-thinking-maas: retires 2026-10-21 (in 27 days) inferred | 2026-09-24 |
| availability | vertex_ai-deepseek_models | vertex_ai/deepseek-ai/deepseek-r1-0528-maas: retires 2026-10-21 (in 27 days) +2 aliases inferred | 2026-09-24 |
| availability | vertex_ai-zai_models | vertex_ai/zai-org/glm-4.7-maas: deprecation announced for 2026-10-21 — in 27 days +1 alias inferred | 2026-09-24 |
| availability | fireworks_ai | fireworks_ai/accounts/fireworks/models/deepseek-v4-flash-0731: deprecation announced for 2026-09-25 — in 1 days +17 aliases | 2026-09-24 |
| availability | vertex_ai-qwen_models | vertex_ai/qwen/qwen3-235b-a22b-instruct-2507-maas: deprecation announced for 2026-10-21 — in 27 days +3 aliases inferred | 2026-09-24 |
| availability | vertex_ai-deepseek_models | vertex_ai/deepseek-ai/deepseek-r1-0528-maas: deprecation announced for 2026-10-21 — in 27 days +2 aliases inferred | 2026-09-24 |
| availability | vertex_ai | vertex_ai/deepseek-ai/deepseek-ocr-maas: retires 2026-10-21 (in 27 days) inferred | 2026-09-24 |
| availability | vertex_ai-openai_models | vertex_ai/openai/gpt-oss-20b-maas: deprecation announced for 2026-10-21 — in 27 days inferred | 2026-09-24 |
| availability | fireworks_ai | fireworks_ai/accounts/fireworks/models/deepseek-v4-flash-0731: retires 2026-09-25 (in 1 days) +17 aliases | 2026-09-24 |
| availability | vertex_ai-minimax_models | vertex_ai/minimaxai/minimax-m2-maas: deprecation announced for 2026-10-21 — in 27 days inferred | 2026-09-24 |
| availability | vertex_ai-minimax_models | vertex_ai/minimaxai/minimax-m2-maas: retires 2026-10-21 (in 27 days) inferred | 2026-09-24 |
| availability | vertex_ai-zai_models | vertex_ai/zai-org/glm-4.7-maas: retires 2026-10-21 (in 27 days) +1 alias inferred | 2026-09-24 |
| availability | vertex_ai-moonshot_models | vertex_ai/moonshotai/kimi-k2-thinking-maas: deprecation announced for 2026-10-21 — in 27 days inferred | 2026-09-24 |
| availability | vertex_ai-llama_models | vertex_ai/meta/llama-3.3-70b-instruct-maas: deprecation announced for 2026-10-21 — in 27 days inferred | 2026-09-24 |
| availability | vertex_ai-llama_models | vertex_ai/meta/llama-3.3-70b-instruct-maas: retires 2026-10-21 (in 27 days) inferred | 2026-09-24 |
| availability | vertex_ai | vertex_ai/deepseek-ai/deepseek-ocr-maas: deprecation announced for 2026-10-21 — in 27 days inferred | 2026-09-24 |
| availability | vertex_ai-openai_models | vertex_ai/openai/gpt-oss-20b-maas: retires 2026-10-21 (in 27 days) inferred | 2026-09-24 |
| availability | vertex_ai-qwen_models | vertex_ai/qwen/qwen3-235b-a22b-instruct-2507-maas: retires 2026-10-21 (in 27 days) +3 aliases inferred | 2026-09-24 |
| availability | fireworks_ai | fireworks_ai/accounts/fireworks/models/deepseek-v4-pro: retires 2026-08-27 (20 days ago) +7 aliases | 2026-09-24 |
| availability | fireworks_ai | fireworks_ai/accounts/fireworks/models/deepseek-v4-pro: deprecation announced for 2026-08-27 (date already passed) +7 aliases | 2026-09-24 |
| availability | openai | openai/sora-2-pro-high-res: retires 2026-09-24 (in 0 days) +8 aliases | 2026-09-24 |
| availability | openrouter | openrouter/nex-agi/nex-n2.5-mini:free: retires 2026-09-25 (in 3 days) +1 alias | 2026-09-22 |
| availability | openrouter | openrouter/nex-agi/nex-n2.5-mini:free: deprecation announced for 2026-09-25 — in 3 days +1 alias | 2026-09-22 |
| availability | nex-agi | nex-agi/nex-n2.5-mini:free: retirement announced for 2026-09-25 — in 3 days +1 alias | 2026-09-22 |
| availability | nex-agi | nex-agi/nex-n2.5-mini:free: retires 2026-09-25 (in 3 days) +1 alias | 2026-09-22 |
| availability | azure_ai | azure_ai/MAI-Image-2.5-Flash: retires 2026-10-01 (in 28 days) +2 aliases inferred | 2026-09-21 |
| availability | azure_ai | azure_ai/MAI-Image-2.5-Flash: deprecation announced for 2026-10-01 — in 28 days +2 aliases inferred | 2026-09-21 |
| availability | groq | groq/qwen/qwen3.6-27b: retires 2026-09-14 (7 days ago) | 2026-09-21 |
| availability | groq | groq/qwen/qwen3.6-27b: deprecation announced for 2026-09-14 (date already passed) | 2026-09-21 |
| availability | openrouter | openrouter/baidu/ernie-4.5-vl-424b-a47b: deprecation announced for 2026-10-08 — in 18 days +1 alias | 2026-09-20 |
| availability | openrouter | openrouter/baidu/ernie-4.5-vl-424b-a47b: retires 2026-10-08 (in 18 days) +1 alias | 2026-09-20 |
| availability | openrouter | openrouter/deepseek/deepseek-r1-distill-llama-70b: deprecation announced for 2026-09-28 — in 8 days +3 aliases | 2026-09-20 |
| availability | openrouter | openrouter/deepseek/deepseek-r1-distill-llama-70b: retires 2026-09-28 (in 8 days) +3 aliases | 2026-09-20 |
| availability | minimax | minimax/minimax-m2.1: retires 2026-10-08 (in 18 days) | 2026-09-20 |
| availability | baidu | baidu/ernie-4.5-vl-424b-a47b: retirement announced for 2026-10-08 — in 18 days | 2026-09-20 |
| availability | deepseek | deepseek/deepseek-r1-distill-llama-70b: retirement announced for 2026-09-28 — in 8 days +3 aliases | 2026-09-20 |
| availability | deepseek | deepseek/deepseek-r1-distill-llama-70b: retires 2026-09-28 (in 8 days) +3 aliases | 2026-09-20 |
| availability | baidu | baidu/ernie-4.5-vl-424b-a47b: retires 2026-10-08 (in 18 days) | 2026-09-20 |
| availability | minimax | minimax/minimax-m2.1: retirement announced for 2026-10-08 — in 18 days | 2026-09-20 |
| availability | databricks | databricks/databricks-gemini-2-5-flash: deprecation announced for 2026-10-02 — in 26 days +1 alias inferred | 2026-09-19 |
| availability | databricks | databricks/databricks-gemini-2-5-flash: retires 2026-10-02 (in 26 days) +1 alias inferred | 2026-09-19 |
| availability | openrouter | openrouter/dots-studio/dots-3-note-preview:free: retires 2026-09-30 (in 12 days) | 2026-09-18 |
| availability | azure | azure/eu/gpt-4o-2024-05-13: retires 2026-10-01 (in 13 days) +7 aliases | 2026-09-18 |
| availability | azure | azure/eu/gpt-4.1-nano: retires 2026-10-14 (in 26 days) +4 aliases | 2026-09-18 |
| availability | azure | azure/eu/gpt-4.1-nano: deprecation announced for 2026-10-14 — in 26 days +2 aliases | 2026-09-18 |
| availability | azure | azure/eu/gpt-4o-2024-05-13: deprecation announced for 2026-10-01 — in 13 days +6 aliases | 2026-09-18 |
| availability | wandb | wandb/MiniMaxAI/MiniMax-M2.5: deprecation announced for 2026-08-25 (date already passed) +1 alias | 2026-09-17 |
| availability | wandb | wandb/JetBrains/Mellum2-12B-A2.5B-Instruct: retires 2026-10-05 (in 18 days) +8 aliases | 2026-09-17 |
| availability | wandb | wandb/JetBrains/Mellum2-12B-A2.5B-Instruct: deprecation announced for 2026-10-05 — in 18 days +8 aliases | 2026-09-17 |
| availability | together_ai | together_ai/deepseek-ai/DeepSeek-V4-Flash-0731: retires 2026-09-29 (in 12 days) +1 alias | 2026-09-17 |
| availability | together_ai | together_ai/deepseek-ai/DeepSeek-V4-Flash-0731: deprecation announced for 2026-09-29 — in 12 days +1 alias | 2026-09-17 |
| availability | together_ai | together_ai/google/gemma-4-31B-it: deprecation announced for 2026-09-14 (date already passed) +3 aliases | 2026-09-17 |
| availability | wandb | wandb/MiniMaxAI/MiniMax-M2.5: retires 2026-08-25 (23 days ago) +1 alias | 2026-09-17 |
| availability | together_ai | together_ai/google/gemma-4-31B-it: retires 2026-09-15 (1 days ago) +3 aliases | 2026-09-16 |
| availability | friendliai | friendliai/LGAI-EXAONE/K-EXAONE-2.0-750B-A37B: retires 2026-09-06 (7 days ago) inferred | 2026-09-13 |
| availability | together_ai | together_ai/moonshotai/Kimi-K2.6: retires 2026-08-19 (25 days ago) | 2026-09-13 |
| availability | together_ai | together_ai/google/gemma-4-31B-it: retires 2026-09-14 (in 1 days) +3 aliases | 2026-09-13 |
| availability | openai | gpt-5.4-cyber: retires 2026-10-01 (in 19 days) inferred | 2026-09-12 |
| availability | dots-studio | dots-studio/dots-3-note-preview:free: retirement announced for 2026-09-30 — in 19 days +1 alias inferred | 2026-09-11 |
| availability | scaleway | scaleway/mistralai/pixtral-12b-2409: retires 2026-10-01 (in 21 days) +1 alias | 2026-09-10 |
| availability | scaleway | scaleway/mistralai/pixtral-12b-2409: deprecation announced for 2026-10-01 — in 21 days +1 alias | 2026-09-10 |
| availability | bedrock_converse | us-gov.anthropic.claude-3-haiku-20240307-v1:0: retires 2026-09-10 (in 1 days) inferred | 2026-09-09 |
| availability | azure | azure/gpt-realtime-2: retires 2026-08-31 (6 days ago) inferred | 2026-09-06 |
| availability | z-ai | z-ai/glm-4.7-flash: retirement announced for 2026-09-10 — in 6 days | 2026-09-04 |
| availability | z-ai | z-ai/glm-4.7-flash: retires 2026-09-10 (in 6 days) | 2026-09-04 |