MiMo-V2.6-Pro
Top open-weights model on the Artificial Analysis index, at cents per task; closed leaders still rank well above it.
Should you use it?
Pick it if
Teams who want near-top coding and agent work for cents per task, or an MIT-licensed model to host themselves.
Best for: Cheap coding and agent work, or an MIT model to host yourself (Chat models Top 20)
The short version
- What this is
- Xiaomi's flagship open model: a 1.02T-parameter mixture-of-experts model (42B active), MIT licensed, with 1M context.
- Why it matters
- It is the top open-weights model on the Artificial Analysis index and one of the cheapest per task, so it is a strong low-cost default.
- Rank 19 of 142 on arena.ai WebDev, 19 of 52 on Agent, 13 of 364 on Expert s1 s3 s6.
- About $0.13 per Artificial Analysis task, against $2.00 for Kimi K3 s5.
- MIT weights, but self-hosting needs a multi-GPU cluster s8.
Good at
Our take
MiMo-V2.6-Pro is the highest-scoring open-weights model on the Artificial Analysis index s5. It sits around rank 11 to 19 on the coding, agent and expert boards s1 s2 s3 s6, which is the same band as Kimi K3 and above GLM-5.3, and its cost per task is a small fraction of theirs s5. arena.ai marks it on the Agent price-versus-score frontier s4.
The weights are MIT licensed s8, but at 1.02T parameters they are for teams with a GPU cluster. Most people will use it through an API, so note that Xiaomi stores platform data in China s7. Xiaomi's launch post lists $0.435 input and $0.87 output per 1M tokens s9; per task, Artificial Analysis puts it at about $0.13 s5.
Scores by job
A score from 0 to 100 per job, against everything available today. Overall 73 across 2 jobs, the average of the job scores.
New to this? What the scores and labels mean
- 0 to 100
- Our score for one job, like writing or coding, compared with everything else out today. 100 is the best available today. For AI models we build it from public data, such as Arena's blind votes and Epoch AI's independent tests; apps and tools are scored on our own checklist. The same tool can score differently for different jobs, and where two likely ranges overlap, we call it a tie.
- Our verdict
- Use it: a top choice for most people. Worth trying: good, and worth a go. Depends: good for some people or one kind of job. Wait: too early to tell, or not out for everyone yet. Skip: something else does the job better.
- First look
- Just released. We say what it is and whether to bother, and give it a score once public results exist.
- No score yet
- None of the public sources we use covers it yet, or the only evidence is from a source whose numbers we can't use, so we describe it in words instead.
- Best buy
- Nothing cheaper scores as well for that job.
- Best free
- The best score for that job that costs nothing, through a free plan or by running on your own computer.
- Pay as you go
- No monthly plan: you pay for each use, usually through a developer account.
- Checked
- The day we last confirmed the score and the price against the sources. Prices change often, so check the date.
- AI model
- The engine underneath, like GPT or Claude. You use it through an app or, if you code, directly.
- Open source
- Free to download and run on your own computer, so nothing leaves your machine.
#3 of the 15 we have scored for this job.
From blind votes (Arena) · Arena data, CC BY 4.0
2 sourcesData read 20 Oct
#16 of the 21 we have scored for this job.
From blind votes (Arena) · Arena data, CC BY 4.0
1 source so farData read 20 Oct
#13 in Arena's blind votes for agents. Those votes give a rank but no rating, so there is no score yet.
#13 in Arena's blind votes for agents (rank only, no score yet) · Arena data, CC BY 4.0
2 sourcesData read 20 Oct
Alternatives
Better or cheaper for agents, from our scores, then the ones we'd also look at. The same kind of product comes first.
- Claude Opus 5.5BetterText model#1$20 a monthRanked #1 on Arena's Agent board in agents, against #13 here; $20 a month. Side by side
- Claude Sonnet 5.5Text model#5$20 a monthMuch higher on coding and agent boards, at a far higher price per task. Side by side
- Kimi K3Text modelNot reviewed yetOpen weights and a few places higher on WebDev, Agent and Expert, at about $2.00 per index task against $0.13 s1 s3 s6 s5.
- MiMo-V2.6-FlashText modelNot reviewed yetCheaper sibling at $0.14 / $0.28 per 1M tokens s9 and rank 16 on WebDev, but rank 32 on Agent and lower on Text Coding s1 s2 s3.
Value for money
Where it sits in agents: higher is a better score, further left is cheaper. Green dots are the Best buys.
Evidence
Everything behind the call and the scores, folded so the page stays short. Spotted a wrong price, score or link? Report a mistake.
How we scored this
The research and coding scores are our own, built from public data: blind votes on Arena and, where they cover the job, Epoch AI's independent tests. How we combine them is our method and we don't publish it. For agents, Arena's blind votes give a rank but no rating, so we show the rank and no score yet.
Public leaderboardIndependent test
Reviews are drafted with help from AI models, including Anthropic's Claude. Model scores come only from outside leaderboards and independent evaluations, by a written rule.
No one paid for this review or gave us free access, and we earn nothing if you sign up. How we score
Sources 9
- s1arena.ai Code Arena: WebDev leaderboard
- s2arena.ai Text: Coding leaderboard
- s3arena.ai Agent leaderboard
- s4arena.ai Agent Pareto and cost per task
- s5Artificial Analysis LLM leaderboard
- s6arena.ai Text: Expert leaderboard
- s7Xiaomi MiMo Platform privacy policy
- s9Xiaomi MiMo-V2.6 launch post (prices)
- s8MiMo-V2.6-Pro-RL model card (Hugging Face)
Something here out of date or wrong? Report a mistake and we will check it against the source.
History 6 changes
Every change to a rating, the price or our call, with the reason. Scores are against what exists on the day, so they drift down as better things ship. Older entries come from before we moved to scores.
- New: Coding ratedNew: Agents ratedNew: Research ratedReviewed: coding, agents, research.
- Audit: removed the claim that it is the top open model for coding and agents. On WebDev, Kimi K3, MiMo-V2.6-Flash, Qwen3.8-Flash-Next and Hy4 preview rank above it, and on Agent, DeepSeek-V4.1-Flash, Kimi K3 and Hy4 preview do. It leads open-weights models only on the Artificial Analysis index. Added Kimi K3 and MiMo-V2.6-Flash as alternatives. Ratings unchanged.
- Audit: confirmed the tie rule. It is tied with Grok 4.7 at 46 on the Artificial Analysis index, so both take rank 9. Still the same, no rating change.
- Added Xiaomi's token prices from its launch post. Ratings unchanged.
- Audit: skip line now names GPT-6 Astra, the best non-Anthropic option, next to Claude Sonnet 5.5. Ranks re-confirmed. Ratings unchanged.
- Plainer how-to-get-it wording; no rating change.
Use it well 4 guides on this site
Pages on this site that help with the jobs above.