MiMo-V2.6-Flash-RL-MLX-4bit-MTP

<p align="center"> <a href="https://tensorfold.dev"> <img src="https://huggingface.co/spaces/TensorFold/README/resolve/main/tensorfold-logo.png" alt="TensorFold" width="160"> </a> </p> <div align="center"> <picture> <source srcset="https://github.com/XiaomiMiMo/MiMo/raw/main/figures/Xiaomi_MiMo_darkmode.png?raw=true" media="(prefers-color-scheme: dark)"> <img src="https://github.com/XiaomiMiMo/MiMo/raw/main/figures/Xiaomi_MiMo.png?raw=true" width="48%" alt="Xiaomi MiMo"> </picture> </div> <h1 align="center">MiMo-V2.6-Flash-RL MLX 4-bit MTP</h1> <p align="center"> A tested Apple Silicon conversion of <a href="https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL">XiaomiMiMo/MiMo-V2.6-Flash-RL</a>, published by <a href="https://huggingface.co/TensorFold">TensorFold</a>. </p>

What is in this release

This is the text backbone of MiMo-V2.6-Flash-RL converted for MLX. The dense projections use 4-bit affine quantization with group size 64, while the model's native MXFP4 MoE experts remain in their original group-size-32 format. The resulting main model averages 4.257 bits per weight.

The checkpoint includes its native three-layer MTP payload in mtp/model_mtp.safetensors, converted to 4-bit affine weights. It also carries the upstream five-layer DFlash drafter, vision encoder, audio encoder, and audio tokenizer so those assets do not need a second download.

Current oMLX and MLX text generation run the target model correctly but do not automatically execute MiMo's MTP, DFlash, vision, or audio paths. The speed figures below are serial text decode measurements. The auxiliary tensors and configs are packaged for MiMo-aware runtimes and ongoing MLX integration, not advertised as working oMLX controls.

Measured on Apple Silicon

Tested on a 256 GB M3 Ultra Mac Studio with oMLX 0.7.0.dev2, MLX 0.32.2, and mlx-lm 0.31.3.

The model produced correct arithmetic, a clear factual explanation, and coherent Python in repeated smoke tests. A 256 GB Mac is recommended so there is room for the model, KV cache, and the rest of the system.

Run with mlx-lm

pip install -U "mlx-lm>=0.31.3"

python -m mlx_lm generate \
  --model TensorFold/MiMo-V2.6-Flash-RL-MLX-4bit-MTP \
  --prompt "Write a Python function that checks whether an integer is prime." \
  --max-tokens 256 \
  --temp 0.6

The upstream tokenizer chat template is included. In oMLX, download or select TensorFold/MiMo-V2.6-Flash-RL-MLX-4bit-MTP as an MLX model.

Quantization notes

MiMo-V2.6 stores fused attention tensors in checkpoint tensor-parallel order and pads the FP8 scale grid separately for each shard. This conversion reconstructs those shards before quantization. Skipping that step produces a model that loads but returns broken output.

The quantized MTP payload lives in its own mtp/ directory with a manifest describing its tensors. It is not a standalone drafter for mlx_lm.generate --draft-model today. The upstream DFlash payload retains its trained mask embedding and corrected JSON config; an MLX smoke test matched serial greedy output, but it did not beat serial decode in the current experimental runtime.

The vision and audio tensors are kept outside the root text-model index so mlx_lm and oMLX continue to load the tested text checkpoint unchanged. omnimodal/manifest.json records every auxiliary path and its upstream source.

About MiMo-V2.6-Flash-RL

Xiaomi describes MiMo-V2.6-Flash-RL as a sparse 309B-parameter MoE with 15B active parameters, 48 transformer layers, 256 routed experts, and eight active experts per token. The full upstream release supports a one-million-token context and omnimodal inputs. See the original model card for architecture details, evaluations, deployment recipes, intended use, and limitations.

License and credit

The upstream model is released under the MIT license. All model architecture, training, tokenizer work, and original branding belong to the Xiaomi MiMo team. This repository contains a community MLX conversion and measured Apple Silicon results.

@misc{mimo2026v26flash,
  title={MiMo-V2.6-Flash-RL},
  author={{Xiaomi MiMo Team}},
  year={2026},
  howpublished={\url{https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL}},
}

Follow TensorFold for new Apple Silicon releases and fixes.