We’ve added MiMo 2.6 Flash and MiMo 2.6 Pro to OpenCheese.
Both models support text, images, video, and audio, offer a context window of up to one million tokens, and are designed for coding, automation, research, and long-running AI agents.
Which model should you choose?
MiMo 2.6 Flash is the efficiency-focused option. Its sparse Mixture-of-Experts architecture contains 309B total parameters but activates only 15B per token, providing strong capabilities with lower latency and cost.
MiMo 2.6 Pro is Xiaomi’s flagship for demanding workloads. It scales to 1.02T total parameters with 42B active and is intended for complex coding, professional automation, and long-horizon agent tasks.
Benchmark comparison
| Benchmark | MiMo 2.6 Pro | MiMo 2.6 Flash | Claude Opus 5 | GPT-5.6 Sol | Claude Fable 5 |
|---|---|---|---|---|---|
| DeepSWE v1.1 | 71.9 | 67.9 | 74.0 | 73.0 | 70.0 |
| AutomationBench v1.0.6 | 53.1 | 52.3 | 50.3 | 45.8 | 46.2 |
| Terminal Bench 2.1 | 89.9 | 87.6 | 89.1 | 88.8 | 84.3 |
| MiMo VisualCoding | 72.3 | 71.5 | 70.0 | 73.4 | 69.1 |
These are Xiaomi-reported results, so real-world performance may vary by prompt, tools, and workflow. Still, they show that Flash offers an excellent balance of capability and efficiency, while Pro competes closely with leading frontier models.
You can try MiMo 2.6 Flash and MiMo 2.6 Pro at opencheese.dev right now!