From 06ff8571aa90bc92177660583b93bec05292788b Mon Sep 17 00:00:00 2001 From: Stephane Date: Fri, 5 Jun 2026 14:34:22 +0200 Subject: [PATCH] fix(model_prices): correct max_input_tokens for minimax/MiniMax-M3 (512K -> 1M) MiniMax-M3 ships a 1,048,576-token (1M) context window per MiniMax's official specs. The 512K figure is MiniMax's long-context billing threshold (requests <= 512K input are billed at the base rate), not the model's context window. This also makes the entry consistent with the sibling MiniMax-M2.x entries, which all use 1000000. Follow-up to #29412. Co-Authored-By: Claude Opus 4.8 (1M context) --- litellm/model_prices_and_context_window_backup.json | 2 +- model_prices_and_context_window.json | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/litellm/model_prices_and_context_window_backup.json b/litellm/model_prices_and_context_window_backup.json index da8ac7404ea5..62cc6097fda6 100644 --- a/litellm/model_prices_and_context_window_backup.json +++ b/litellm/model_prices_and_context_window_backup.json @@ -24132,7 +24132,7 @@ "supports_reasoning": true, "supports_system_messages": true, "supports_vision": true, - "max_input_tokens": 512000, + "max_input_tokens": 1048576, "max_output_tokens": 128000 }, "mistral.devstral-2-123b": { diff --git a/model_prices_and_context_window.json b/model_prices_and_context_window.json index 5699a38db319..6db2b867f0d4 100644 --- a/model_prices_and_context_window.json +++ b/model_prices_and_context_window.json @@ -24132,7 +24132,7 @@ "supports_reasoning": true, "supports_system_messages": true, "supports_vision": true, - "max_input_tokens": 512000, + "max_input_tokens": 1048576, "max_output_tokens": 128000 }, "mistral.devstral-2-123b": {