Repository navigation
Add Minimax-M3 - #2621
Add Minimax-M3#2621
Conversation
There was a problem hiding this comment.
Pull request overview
Note
Copilot was unable to run its full agentic suite in this review.
Adds a new NVIDIA provider model configuration for minimaxai/minimax-m3.
Changes:
- Introduces a new TOML model manifest for
minimax-m3 - Defines structured output support, pricing placeholders, token limits, and supported modalities
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| input = 0.0 | ||
| output = 0.0 | ||
| cache_read = 0.0 |
There was a problem hiding this comment.
The model IS free at this provider, so it is correct
| context = 1000000 | ||
| input = 1000000 | ||
| output = 131072 |
|
go merge this please |
|
AUTOMATED REVIEW: Blocking merge: CI fails because the |
Done |
| @@ -0,0 +1,16 @@ | |||
| base_model = "minimax/MiniMax-M3" | |||
| reasoning = true | |||
|
|
|||
There was a problem hiding this comment.
missing reasoning_options
There was a problem hiding this comment.
@rekram1-node could you clarify the expected format for reasoning_options? Should it be [[reasoning_options]] with type/values, or a simple array? And what are the correct values for the NVIDIA MiniMax-M3 toggle — "enabled"/"disabled", true/false, or something else?
Removed reasoning_options from minimax-m3.toml
|
@rekram1-node removed reasoning_options as it caused a schema invalid_type error — the validator rejected the format. CI passes now. Could you clarify what exact syntax you need for that field so I can add it correctly? |
|
Completed and added missing reasoning_options @rekram1-node |
|
ur not using base_model syntax properly |
Adds NVIDIA's
minimaxai/minimax-m3model using the canonical MiniMax M3 metadata plus NVIDIA-specific limits.Changes
providers/nvidia/models/minimaxai/minimax-m3.toml.base_model = "minimax/MiniMax-M3".max_tokensoutput limit of16_384.chat_template_kwargs.thinking_mode = "enabled" | "disabled" | "adaptive"for this model.Evidence
minimaxai/minimax-m3: https://integrate.api.nvidia.com/v1/modelsmax_tokensrange1..16_384andchat_template_kwargs.thinking_mode: https://docs.api.nvidia.com/nim/reference/minimaxai-minimax-m3-inferValidation
bun validategit diff --check