$ diff mimo-v2-6-pro hy4-preview
MiMo-V2.6-Pro vs Hy4 preview
Values come from MiMo-V2.6-Pro and Hy4 preview, where each is cited to its source. This page states no benchmark result and ranks neither model — it puts two published specifications next to each other. Where a lab has not published a figure, the row says so rather than guessing.
XiaomiMiMo-V2.6-Pro
- context
- 1M
- weights
- open
- $/M in
- $0.435
- $/M out
- $0.87
- context
- 1M
- weights
- open
- $/M in
- $0.834
- $/M out
- $2.501
What actually differs
- Context window
- Identical — both accept 1M tokens, so context length is not a reason to pick either.
- Input price
- MiMo-V2.6-Pro at $0.435/M against $0.834/M — 1.9× cheaper to feed.
- Output price
- MiMo-V2.6-Pro at $0.87/M against $2.501/M — 2.9× cheaper to generate. Output dominates the bill on most agentic workloads, where the model writes far more than it reads.
- Weights
- Both publish weights — MiMo-V2.6-Pro under MIT (open weights), Hy4 preview under Apache-2.0. Check the licences rather than assuming they permit the same commercial use.
- Recency
- MiMo-V2.6-Pro shipped 25 days after Hy4 preview (2026-09-22 vs 2026-08-28).
Full spec
| Attribute | MiMo-V2.6-Pro | Hy4 preview |
|---|---|---|
| Developer | Xiaomi | Tencent |
| Released | 2026-09-22 | 2026-08-28 |
| Context window | 1M tokens | over 1,000,000 |
| Pricing | no vendor list price — weights are MIT and self-hostable; routed providers reported at $0.435/M input · $0.87/M output | $0.834/M input · $2.501/M output |
| License | MIT (open weights) | Apache-2.0 |
| Availability | Hugging Face as `XiaomiMiMo/MiMo-V2.6-Pro-RL`; hosted via third-party routers | Hugging Face (BF16 + FP8), ModelScope, GitCode, CNB, Tencent Cloud TokenHub, OpenRouter, WorkBuddy, CodeBuddy, Yuanbao, ima |