Runs on ≤8 GB · Hugging Face ↗
Not independently scored by Artificial Analysis — its base model Qwen/Qwen3.5-9B is the closest proxy
Xiaomi distils its MiMo V2.6 into a 9B on Qwen3.5 (MIT), with an official GGUF from ggml. The big model of the same family is what Xiaomi sells inside its token plan.
BigCodeBench-Hard pass@1 18% (26/148), via official protocol · llama.cpp Q4_K_M · one request at a time · a shared cluster GPU · mean of 6 runs, range 0.0 pp. The brutal counterpart to HumanEval — where HumanEval saturates near the top, BCB-Hard spreads the field, so this is the number that actually separates coding ability. repeats itself until it runs out; 3 of 148 answers hit the token limit (effective ceiling 98%)
Our own runs on this model. Each number carries how many problems it is over and how many times we ran it — these are not comparable with each other, and none of them is on the same scale as the third-party indices on the leaderboard.
| test | score | over | conditions |
|---|---|---|---|
| BigCodeBench-Hard | 17.6% | n=148 · 6 run(s) | official protocol · llama.cpp Q4_K_M · one request at a time · a shared cluster GPU · mean of 6 runs, range 0. |
Real GGUF file sizes = weight VRAM. Add the KV cache for your context (≈1 GB at 32K, fp16). Estimated from the original model config, not from the GGUF we serve. Less reliable than the rest of this page.
| Quantization | Size |
|---|---|
| Q8_0 | 9.5 GB |
17.6% over three runs, repeated twice a week apart with the same number. A 9B that scores like the 14B baselines, from an official GGUF.