MiMo-V2-Flash (Feb 2026) vs MiMo v2.6 Flash
Humanity's Last Exam Humanity's Last Exam: Very hard expert questions across many fields, written to stump today’s models
MiMo-V2-Flash (Feb 2026)22.1
MiMo v2.6 Flash35.1
Arena Text · overall Arena Text · overall: Human pairwise preferences in the stated text vote pool
MiMo-V2-Flash (Feb 2026)1,386
MiMo v2.6 Flash1,451
Arena Text · Chinese Arena Text · chinese: Human pairwise preferences in the stated text vote pool
MiMo-V2-Flash (Feb 2026)1,428.6
MiMo v2.6 Flash1,513.3
Arena Text · English Arena Text · english: Human pairwise preferences in the stated text vote pool
MiMo-V2-Flash (Feb 2026)1,411.7
MiMo v2.6 Flash1,477.4
Arena Text · French Arena Text · french: Human pairwise preferences in the stated text vote pool
MiMo-V2-Flash (Feb 2026)1,428.8
MiMo v2.6 Flash1,471.6
Arena Text · German Arena Text · german: Human pairwise preferences in the stated text vote pool
MiMo-V2-Flash (Feb 2026)1,395
MiMo v2.6 Flash1,464.3
Arena Text · instruction following Arena Text · instruction following: Human pairwise preferences in the stated text vote pool
MiMo-V2-Flash (Feb 2026)1,372.6
MiMo v2.6 Flash1,456.1
Arena Text · Korean Arena Text · korean: Human pairwise preferences in the stated text vote pool
MiMo-V2-Flash (Feb 2026)1,287.1
MiMo v2.6 Flash1,418.2
Arena Text · non-English Arena Text · non english: Human pairwise preferences in the stated text vote pool
MiMo-V2-Flash (Feb 2026)1,373.4
MiMo v2.6 Flash1,432.8
Arena Text · Polish Arena Text · polish: Human pairwise preferences in the stated text vote pool
MiMo-V2-Flash (Feb 2026)1,392.9
MiMo v2.6 Flash1,469.7
Arena Text · Russian Arena Text · russian: Human pairwise preferences in the stated text vote pool
MiMo-V2-Flash (Feb 2026)1,373
MiMo v2.6 Flash1,409.8
Arena Text · Spanish Arena Text · spanish: Human pairwise preferences in the stated text vote pool
MiMo-V2-Flash (Feb 2026)1,393.6
MiMo v2.6 Flash1,455.5
AA Omniscience Index AA Omniscience Index: Scores factual knowledge while penalizing incorrect answers to encourage admitting uncertainty
MiMo-V2-Flash (Feb 2026)-19.1
MiMo v2.6 Flash-12.7
CritPt (physics) CritPt (physics): Research-level physics problems that need many careful reasoning steps
MiMo-V2-Flash (Feb 2026)2.9
MiMo v2.6 Flash12
AA Intelligence Index 4.3.2 AA Intelligence Index 4.3.2: Artificial Analysis combines tests of knowledge, reasoning, coding and tool use into one score
MiMo-V2-Flash (Feb 2026)22.4
MiMo v2.6 Flash37.9
AA Omniscience · Accuracy AA Omniscience · Accuracy: Measures correct answers on the AA Omniscience factual knowledge test
MiMo-V2-Flash (Feb 2026)19.8
MiMo v2.6 Flash27
AA Omniscience · Hallucination Rate AA Omniscience · Hallucination Rate: Measures how often attempted answers are incorrect on the AA Omniscience factual knowledge test · lower is better
MiMo-V2-Flash (Feb 2026)48.4
MiMo v2.6 Flash54.4
AA-LCR (long context) 1.1 AA-LCR (long context) 1.1: Questions that need reading and connecting facts across very long documents (around 100K tokens)
MiMo-V2-Flash (Feb 2026)71.3
MiMo v2.6 Flash74.3
Arena Text · Japanese Arena Text · japanese: Human pairwise preferences in the stated text vote pool
MiMo-V2-Flash (Feb 2026)—
MiMo v2.6 Flash1,353.8
GDP.pdf · AA all-pass GDP.pdf · AA all-pass: Models reason over professional PDF documents; an attempt passes only when every atomic criterion passes
MiMo-V2-Flash (Feb 2026)—
MiMo v2.6 Flash9
IFBench (instructions) IFBench (instructions): Checks whether the model follows precise formatting and content instructions
MiMo-V2-Flash (Feb 2026)71.8
MiMo v2.6 Flash—
Vals MedCode Vals MedCode: Medical coding accuracy on patient encounters
MiMo-V2-Flash (Feb 2026)—
MiMo v2.6 Flash41.1
Vals MedScribe Vals MedScribe: Accuracy of clinical notes produced from consultation transcripts
MiMo-V2-Flash (Feb 2026)—
MiMo v2.6 Flash85.3
Vals Legal Research Bench Vals Legal Research Bench: Grounded legal research; a task passes only if all criteria pass
MiMo-V2-Flash (Feb 2026)—
MiMo v2.6 Flash38
Vals Science Index Vals Science Index: Scientific problem solving across the index’s component evaluations
MiMo-V2-Flash (Feb 2026)—
MiMo v2.6 Flash31.7
| Benchmark | MiMo-V2-Flash (Feb 2026) | MiMo v2.6 Flash |
|---|---|---|
| Humanity's Last Exam | 22.1 | 35.1 |
| Arena Text · overall | 1,386 | 1,451 |
| Arena Text · Chinese | 1,428.6 | 1,513.3 |
| Arena Text · English | 1,411.7 | 1,477.4 |
| Arena Text · French | 1,428.8 | 1,471.6 |
| Arena Text · German | 1,395 | 1,464.3 |
| Arena Text · instruction following | 1,372.6 | 1,456.1 |
| Arena Text · Korean | 1,287.1 | 1,418.2 |
| Arena Text · non-English | 1,373.4 | 1,432.8 |
| Arena Text · Polish | 1,392.9 | 1,469.7 |
| Arena Text · Russian | 1,373 | 1,409.8 |
| Arena Text · Spanish | 1,393.6 | 1,455.5 |
| AA Omniscience Index | -19.1 | -12.7 |
| CritPt (physics) | 2.9 | 12 |
| AA Intelligence Index 4.3.2 | 22.4 | 37.9 |
| AA Omniscience · Accuracy | 19.8 | 27 |
| AA Omniscience · Hallucination Rate | 48.4 | 54.4 |
| AA-LCR (long context) 1.1 | 71.3 | 74.3 |
| Arena Text · Japanese | — | 1,353.8 |
| GDP.pdf · AA all-pass | — | 9 |
| IFBench (instructions) | 71.8 | — |
| Vals MedCode | — | 41.1 |
| Vals MedScribe | — | 85.3 |
| Vals Legal Research Bench | — | 38 |
| Vals Science Index | — | 31.7 |
Specs
| Spec | MiMo-V2-Flash (Feb 2026) | MiMo v2.6 Flash |
|---|---|---|
| Input, per 1M | — | $0.14 |
| Output, per 1M | — | $0.28 |
| Cost per task? | — | $0.062 |
| Speed, tok/s | — | 41 |
| Context | 256K | 1M |
| Weights | Open | Open |
| Released | Dec 2025 | Sep 2026 |
| Memory (4-bit) | — | 186 GB |
| Licence | — | MIT |
| Commercial use | — | Allowed |
| See images | ? | ✓ |
| Read PDFs | ? | — |
| Hear audio | ? | ✓ |
| Watch video | ? | ✓ |
| Call tools | ? | ✓ |
| Force a tool call | ? | Partial |
| Return strict JSON | ? | ✓ |
| Search the web | ? | ? |
| Run code | ? | ? |
| Use a computer | ? | ? |
What breaks when you switch
MiMo-V2-Flash (Feb 2026) → MiMo v2.6 Flash
Pass back reasoning_content›
In deep-thinking multi-turn tool conversations, return the full reasoning_content in later rounds with tool calls or receive HTTP 400.
Forced recommended defaults›
Deep thinking overrides supplied temperature and top_p with 1.0 and 0.95.
Avoid answer truncation›
max_completion_tokens includes thinking and the final answer; reserve enough tokens to avoid truncating the answer.
MOPD checkpoints address tool-call repetition›
Use the MOPD checkpoints for the tool-call repetition fix; the API has served the updated models since September 25 at 06:00 UTC+8 under unchanged names.
MiMo v2.6 Flash → MiMo-V2-Flash (Feb 2026)