Xiaomi's MiMo-V2.6-Pro debuts as top open-weights model with ~$2.6m RL training bill
andersonbcdefg · x · 2026-09-22
Per Artificial Analysis, Xiaomi's open-weights MiMo-V2.6-Pro debuts as the top open model on its Intelligence Index (46), up sharply from MiMo-V2.5-Pro's 26. The MoE model has 1.02T total / 42B active parameters and keeps pricing at $0.435 per 1M input tokens (99% cache discount) and $0.87 per 1M output — $0.13 per task, landing on the intelligence-vs-cost Pareto frontier.
A widely shared add-on on post-training scaling laws: SoTA was reached with 45k RL tasks, nearly all synthetic, on $2.6m of Chinese compute ($1m US equivalent), suggesting human data is bottlenecks US frontier labs while neolabs prematurely scale compute.
Related event: Xiaomi open-sources MiMo-V2.6, topping open-weight intelligence index(18 posts)→
More from Models
- DFlash2 reportedly chokes on super long prompts when paired with GLM 5.3 Flash — TheZachMueller · 2026-09-22
- Shots fired at DeepSeek: MiMo pioneered MOPD, observers weigh in on V2.6 — teortaxesTex · 2026-09-22
- Jev fails counting r's in strawberry, aces it 168/168 when given letters — BLUECOW009 · 2026-09-22
- Blogger corrects himself: the real surprise is MiMo-2.6, beating Grok 4.7 at much lower cost — kimmonismus · 2026-09-22
- Jev: InstructGPT coauthor Diogo Almeida on System One models for prod, not God — Latent Space · 2026-09-22
- Unverified DataBench Charts Fuel Rumors of OpenAI's Internal Model 'Luna' Ahead of GPT-6 — almmaasoglu · 2026-09-22