MiMo-V2-Flash: benchmark scores, price and ranking
MiMo-V2-Flash from XiaomiMiMo ranks #564 of 743 LLMs on the LLMs Tiger aggregate benchmark score (45.6), based on 5 benchmark results captured up to 2026-10-13.
- Organisation
- XiaomiMiMo
- Type
- Open-weight / local model
- Open weights
- Yes
- Cheapest API price
- $0.110 in / $0.330 out per million tokens (via novita, observed 2026-10-13T20:47Z)
- Aggregate rank
- #564 of 743 (score 45.6)
Relative to other models, MiMo-V2-Flash is strongest on GPQA Diamond (#93 of 275) and weakest on SciCode (#98 of 106).
Across the 5 benchmarks where at least ten models are scored, MiMo-V2-Flash is in the top 10% on 0 and in the top half on 1.
Its cheapest listed input price, $0.110 per million tokens, is lower than that of 79% of the 427 models with a listed price.
MiMo-V2-Flash benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| GPQA Diamond | 83.3% | #93 of 275 | HuggingFace GPQA Diamond leaderboard | 2026-10-13 |
| ALE-Bench | 738 | #42 of 79 | Epoch AI Benchmarking Hub | 2026-10-13 |
| WebDev Arena | 1,331 | #51 of 67 | Epoch AI Benchmarking Hub | 2026-10-13 |
| CritPt | 0.0% | #104 of 115 | Epoch AI Benchmarking Hub | 2026-10-13 |
| SciCode | 25.9% | #98 of 106 | Epoch AI Benchmarking Hub | 2026-10-13 |
About MiMo-V2-Flash
MiMo-V2-Flash is an open-weight model that can run on your own hardware from XiaomiMiMo. The cheapest API listing we track charges $0.110 per million input tokens and $0.330 per million output tokens, through novita. We hold 5 benchmark results for it across 2 areas, and none is in the top tenth of its board.
Where MiMo-V2-Flash stands
On our aggregate score, MiMo-V2-Flash is #564 of 743 ranked models, in the lower half. Its score is 45.6; the scale is centred on 50, so it sits 4.4 points below the typical model. That beats 24% of the ranked field. Among the 3 ranked XiaomiMiMo models it is #3. The next one up is MiMo-7B-Base at 46.0.
MiMo-V2-Flash, benchmark by benchmark
GPQA Diamond puts MiMo-V2-Flash at 83.3%, #93 of 275. That beats 66% of the models tested, but 12.4 points short of GPT-6 Astra (max) (95.8%). The results either side are GLM-5.1-FP8 (83.3%) and Qwen 3.6 Flash (2026-04-16) (83.3%). It covers expert-level science questions. The board median is 71.7%, so MiMo-V2-Flash is 11.6 points above it. Source: HuggingFace GPQA Diamond leaderboard, captured 2026-10-13.
MiMo-V2-Flash lands below the middle on ALE-Bench: 738, #42 of 79, with 52% of the field scoring higher. GPT-6 Astra (max) leads at 2,951; the closest result above is DeepSeek-V3.1-Terminus at 745. It covers AtCoder heuristic-optimisation contests (performance rating). The board median is 764, so MiMo-V2-Flash is 26 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
MiMo-V2-Flash lands below the middle on WebDev Arena: 1,331, #51 of 67, with 75% of the field scoring higher. Claude Opus 5.5 (max) leads at 1,820; the closest result above is GPT-5.1 Codex at 1,336. It covers building web apps, crowd-voted (Elo). The board median is 1,433, so MiMo-V2-Flash is 103 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
CritPt is a weak spot. MiMo-V2-Flash scores 0.0%, #104 of 115, behind 90% of the models measured. The leader, GPT-5.6 Sol (max), reaches 32.3%. It covers unpublished research-level physics problems. The board median is 2.9%, so MiMo-V2-Flash is 2.9 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
SciCode is a weak spot. MiMo-V2-Flash scores 25.9%, #98 of 106, behind 92% of the models measured. The leader, Claude Opus 5.5 (max), reaches 66.9%. It covers scientific research code. The board median is 44.3%, so MiMo-V2-Flash is 18.4 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Strengths and weaknesses
Software engineering: weak, ahead of 26% of the field on average across 3 benchmarks. Its best board here is ALE-Bench (#42 of 79) and its weakest is SciCode (#98 of 106). Reasoning, maths and science: uneven, ahead of 38% of the field on average across 2 benchmarks. Its best board here is GPQA Diamond (#93 of 275) and its weakest is CritPt (#104 of 115).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so MiMo-V2-Flash costs less than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), Qwen2.5-Coder-3B-Instruct ($0.010, 46.5). Similar scores cost more on qwen3-coder-next ($0.120) and Phi-3.5-mini-instruct ($0.130).
Reading these numbers
The 5 scores on this page come from 2 sources: Epoch AI Benchmarking Hub (4), HuggingFace GPQA Diamond leaderboard (1). No official leaderboard supplied any of them, so treat them as third-party measurements. With 5 results the aggregate rank can move a lot when one new score arrives.
Models ranked nearby
- #561 Mixtral-8x7B-Instruct-v0.1 (45.7)
- #562 EXAONE-3.5-2.4B-Instruct (45.7)
- #563 qwen-plus-2025-01-25 (45.7)
- #565 Llama 4 Scout Instruct (45.6)
- #566 MiniMax-M3 (45.6)
- #567 open-mistral-nemo-2407 (45.6)
More from XiaomiMiMo
- MiMo-7B-RL (#364, 48.3)
- MiMo-7B-Base (#541, 46.0)
Compare it on the live leaderboard → · How the aggregate score works