Compare / Meta vs Xiaomi

AI labs compared · November 2026

Meta vs Xiaomi: which AI lab is ahead?

Each lab's best model on sale goes head to head on every kind of work we measure, from our own benchmarks and licensed sources. 15 models in all; a lab is only as good as the model you would actually pick from it.

11 models on sale with our resultsMeta Best: Muse Wonder · newest: Muse Ember
4 models on sale with our resultsXiaomi Best: MiMo-V2.6-Pro · newest: MiMo-V2.6-Pro
Meta wins 3Xiaomi wins 1

Kinds of work led by each lab's best model there, of the 4 both are measured on.

Results as of 19 November 2026 (catalogue release 2026-11-19-3608133e8331). Models on sale means listed on OpenRouter today.

Meta or Xiaomi? The short answer

Meta is ahead: its best models beat Xiaomi's in 3 of the 4 kinds of work both are measured on; Xiaomi leads in 1.

  • Pick Meta for agents and tool use, reasoning and knowledge and writing.
  • Pick Xiaomi for sticking to the facts.
  • Meta's Muse Wonder is #10 on our overall leaderboard; no Xiaomi model has enough of our results to be ranked there yet.
  • Prices start at $0.058 per million tokens for Meta (Llama 3.1 8B Instruct) and $0.18 for Xiaomi (MiMo-V2.6-Flash).

Head to head

52.4Muse Wonder
Best Intelligence Index
46.3MiMo-V2.6-Pro
$8.00Muse Wonder
Flagship price, per million tokens
$0.54MiMo-V2.6-Pro
$0.058Llama 3.1 8B Instruct
Cheapest model, per million tokens
$0.18MiMo-V2.6-Flash
11
Models on sale with our results
4

Each line names the model behind the number. Prices are blended, three input tokens to one output, at the developer's list price where we have read it; see every model's price. Intelligence Index from Artificial Analysis; overall score from our leaderboard.

Which lab is better for what

For each kind of work, each lab's best model there and its rank among every model on sale. Further right is better; the better rank is in green.

Product listingsTurning a sparse product feed and photos into listings that can go live 8 of 35Muse Wonder not measured
Decks from an analysisTurning a finished analysis into a deck you could present as it is 15 of 35Muse Wonder not measured
User surveysPlanning a user survey and reading its results without being misled 3 of 26Muse Wonder not measured
Marketing planningPlanning a year of ad spend without overspending 10 of 35Muse Ember not measured
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 11 of 159Muse Wonder 23 of 159MiMo-V2.6-Pro
Professional workReal tasks from banking, consulting and law, business documents and freelance projects 8 of 51Muse Spark 1.3 not measured
Reasoning and knowledgeHard questions across science, maths and general knowledge 8 of 187Muse Wonder 13 of 187MiMo-V2.6-Pro
WritingWhat people prefer in blind comparisons, and judged writing quality 7 of 169Muse Spark 1.3 23 of 169MiMo-V2.6-Pro
Sticking to the factsSummarising without inventing things, and factual answers 11 of 139Muse Spark 1.3 10 of 139MiMo-V2.6-Pro
Speed under pressureGood decisions against a real clock (fast chess) 17 of 38Muse Wonder not measured

Every model, by intelligence and price

Each dot is a model on sale with an Intelligence Index from Artificial Analysis. Up is smarter; left is cheaper. Hover or tap a dot for its name.

MetaXiaomi

0204060 $0.01$0.1$1$10 Price per million tokens, blended (log scale) Intelligence Index Muse WonderMuse Spark 1.3Muse Spark 1.2Muse Spark 1.1MiMo-V2.6-ProMiMo-V2.6-Flash

On our overall leaderboard

The 3 of their models we rank overall: one score out of 100 across everything we publish (how the score works). Only models with enough of our own results are ranked.

  1. #10Muse Wonder66
  2. #13Muse Ember61
  3. #17Muse Spark 1.358

Benchmark by benchmark

The headline result of every benchmark both labs have models on: each lab's best model and its rank among the models on sale. Green marks the better rank.

MeasureMetaXiaomi
Agents and tool use
Artificial Analysis · Terminal-Bench 4.0% of tasks, higher is better 52.0%Muse Wonder · 9 of 10734.8%MiMo-V2.6-Pro · 17 of 107
Artificial Analysis · Terminal-Bench 2.1% of tasks, higher is better 85.4%Muse Spark 1.3 · 10 of 10765.2%MiMo-V2.5-Pro · 38 of 107
Artificial Analysis · τ-bench banking% of tasks, higher is better 50.5%Muse Spark 1.3 · 2 of 1049.9%MiMo-V2.5-Pro · 71 of 104
Reasoning and knowledge
Artificial Analysis · Artificial Analysis Intelligence Indexindex score, higher is better 52.4Muse Wonder · 6 of 18546.3MiMo-V2.6-Pro · 14 of 185
Artificial Analysis · Humanity's Last Exam% of questions, higher is better 50.1%Muse Wonder · 10 of 18649.4%MiMo-V2.6-Pro · 12 of 186
Artificial Analysis · GPQA Diamond% of questions, higher is better 94.1%Muse Spark 1.3 · 5 of 16886.6%MiMo-V2.5-Pro · 62 of 168
Sticking to the facts
Arena (formerly LMArena) · Overall · Text factualityArena rating, higher is better 1,483Muse Spark 1.2 · 10 of 1091,483MiMo-V2.6-Pro · 9 of 109
Writing
Arena (formerly LMArena) · Overall · TextArena rating, higher is better 1,494Muse Spark 1.3 · 8 of 1531,480MiMo-V2.6-Pro · 21 of 153
UGI Leaderboard · Writing score · UGIscore out of 100, higher is better 26.7Llama 4 Scout · 110 of 12457.4MiMo-V2.5-Pro · 38 of 124

Quick answers

Is Meta or Xiaomi better?

Meta is ahead: its best models beat Xiaomi's in 3 of the 4 kinds of work both are measured on; Xiaomi leads in 1.

What is Meta's best model?

Muse Wonder: #10 on the Spring Prompt overall leaderboard and 52.4 on the Artificial Analysis Intelligence Index.

What is Xiaomi's best model?

MiMo-V2.6-Pro: 46.3 on the Artificial Analysis Intelligence Index.

Which is better for agents and tool use, Meta or Xiaomi?

Meta: its best model there, Muse Wonder, ranks 11 of 159 models on sale, against 23 for Xiaomi's MiMo-V2.6-Pro.

Which is better for reasoning and knowledge, Meta or Xiaomi?

Meta: its best model there, Muse Wonder, ranks 8 of 187 models on sale, against 13 for Xiaomi's MiMo-V2.6-Pro.

Which is better for writing, Meta or Xiaomi?

Meta: its best model there, Muse Spark 1.3, ranks 7 of 169 models on sale, against 23 for Xiaomi's MiMo-V2.6-Pro.

Which is cheaper, Meta or Xiaomi?

Meta's cheapest model with results, Llama 3.1 8B Instruct, costs $0.058 per million tokens (three input to one output), against $0.18 for MiMo-V2.6-Flash from Xiaomi.

Other labs compared

Which model is best for your product?

Benchmarks aren't your task. Run models from Meta, Xiaomi and the rest on your own prompt, and see the cheapest one that passes.