[object Object]

← back to Exo

Add MiniMax M2.5 model cards (#1514)

ce5a65d3b979eb5560be1fe975add239c8921e06 · 2026-02-18 13:11:13 -0800 · Alex Cheema

## Summary
- Adds model cards for MiniMax M2.5 in three quantizations: 4bit (~129
GB), 6bit (~186 GB), 8bit (~243 GB)
- No code changes needed — `MiniMaxM2ForCausalLM` is already in the
tensor parallel whitelist and `MiniMaxShardingStrategy` is already
implemented in `auto_parallel.py`
- Credit to @vskiwi for confirming MiniMax M2.5 works out of the box
with existing code

Closes #1480

## Test plan
- [x] `basedpyright` passes with 0 errors
- [x] `ruff check` passes
- [x] `pytest` passes (260 passed, 1 skipped)
- [ ] Verify MiniMax M2.5 models appear in model selector on dashboard

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: rltakashige <rl.takashige@gmail.com>

Files touched

Diff

commit ce5a65d3b979eb5560be1fe975add239c8921e06
Author: Alex Cheema <41707476+AlexCheema@users.noreply.github.com>
Date:   Wed Feb 18 13:11:13 2026 -0800

    Add MiniMax M2.5 model cards (#1514)
    
    ## Summary
    - Adds model cards for MiniMax M2.5 in three quantizations: 4bit (~129
    GB), 6bit (~186 GB), 8bit (~243 GB)
    - No code changes needed — `MiniMaxM2ForCausalLM` is already in the
    tensor parallel whitelist and `MiniMaxShardingStrategy` is already
    implemented in `auto_parallel.py`
    - Credit to @vskiwi for confirming MiniMax M2.5 works out of the box
    with existing code
    
    Closes #1480
    
    ## Test plan
    - [x] `basedpyright` passes with 0 errors
    - [x] `ruff check` passes
    - [x] `pytest` passes (260 passed, 1 skipped)
    - [ ] Verify MiniMax M2.5 models appear in model selector on dashboard
    
    🤖 Generated with [Claude Code](https://claude.com/claude-code)
    
    Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
    Co-authored-by: rltakashige <rl.takashige@gmail.com>
---
 .../mlx-community--MiniMax-M2.5-4bit.toml                    | 12 ++++++++++++
 .../mlx-community--MiniMax-M2.5-6bit.toml                    | 12 ++++++++++++
 .../mlx-community--MiniMax-M2.5-8bit.toml                    | 12 ++++++++++++
 3 files changed, 36 insertions(+)

diff --git a/resources/inference_model_cards/mlx-community--MiniMax-M2.5-4bit.toml b/resources/inference_model_cards/mlx-community--MiniMax-M2.5-4bit.toml
new file mode 100644
index 00000000..d64f29f9
--- /dev/null
+++ b/resources/inference_model_cards/mlx-community--MiniMax-M2.5-4bit.toml
@@ -0,0 +1,12 @@
+model_id = "mlx-community/MiniMax-M2.5-4bit"
+n_layers = 62
+hidden_size = 3072
+supports_tensor = true
+tasks = ["TextGeneration"]
+family = "minimax"
+quantization = "4bit"
+base_model = "MiniMax M2.5"
+capabilities = ["text", "thinking"]
+
+[storage_size]
+in_bytes = 128666664960
diff --git a/resources/inference_model_cards/mlx-community--MiniMax-M2.5-6bit.toml b/resources/inference_model_cards/mlx-community--MiniMax-M2.5-6bit.toml
new file mode 100644
index 00000000..c92dcd0d
--- /dev/null
+++ b/resources/inference_model_cards/mlx-community--MiniMax-M2.5-6bit.toml
@@ -0,0 +1,12 @@
+model_id = "mlx-community/MiniMax-M2.5-6bit"
+n_layers = 62
+hidden_size = 3072
+supports_tensor = true
+tasks = ["TextGeneration"]
+family = "minimax"
+quantization = "6bit"
+base_model = "MiniMax M2.5"
+capabilities = ["text", "thinking"]
+
+[storage_size]
+in_bytes = 185826705408
diff --git a/resources/inference_model_cards/mlx-community--MiniMax-M2.5-8bit.toml b/resources/inference_model_cards/mlx-community--MiniMax-M2.5-8bit.toml
new file mode 100644
index 00000000..b1abb744
--- /dev/null
+++ b/resources/inference_model_cards/mlx-community--MiniMax-M2.5-8bit.toml
@@ -0,0 +1,12 @@
+model_id = "mlx-community/MiniMax-M2.5-8bit"
+n_layers = 62
+hidden_size = 3072
+supports_tensor = true
+tasks = ["TextGeneration"]
+family = "minimax"
+quantization = "8bit"
+base_model = "MiniMax M2.5"
+capabilities = ["text", "thinking"]
+
+[storage_size]
+in_bytes = 242986745856

← c2f2111b Fix tool calling (#1529)  ·  back to Exo  ·  fix: unblock MpReceiver.close() to prevent shutdown hang (#1 31599254 →