Kimi K2 Thinking vs MiMo V2 Flash
Wins 3 of 7 areas
Coding · Facts · Following instructions
Wins 1 of 7 areas
Math
Kimi K2 Thinking wins more areas, narrowly.MiMo V2 Flash is cheaper and better at math.
Scores updated · 36 tests both models report · How we compare
Where each one wins
Tests won in each of the seven areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- FactsGetting facts right instead of making them up20Kimi K2 Thinking2 of 2 tests
- CodingWriting and fixing software32Kimi K2 Thinking3 of 5 tests
- Following instructionsDoing exactly what it is asked10Kimi K2 Thinking1 of 1 test
- MathCompetition and research-level math01MiMo V2 Flash1 of 1 test
- AgentsCarrying out multi-step tasks on its own11Even1 each
- ReasoningHard problems that need careful thinking11Even1 each · 1 tie
- Long documentsFinding answers in very long texts11Even1 each
Math and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
MiMo V2 Flash costs 87% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Kimi K2 Thinking pulls ahead
- Real work tasks from 44 professionsGDPVal+17.9points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+17.8points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+6.1points ahead
Where MiMo V2 Flash pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+16.8points ahead
- Questions about very long textsLongBench v2+15.5points ahead
- Fixes real GitHub issues in many programming languagesSWE-bench Multilingual+10.6points ahead
Every test, side by side
All 36 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingKimi K2 Thinking
- SWE-bench MultilingualMiMo V2 Flash by 10.661.171.7+10.6
- SWE-bench VerifiedMiMo V2 Flash by 1063.473.4+10
- Terminal-Bench HardKimi K2 Thinking by 3.131.128+3.1
- SciCodeKimi K2 Thinking by 342.439.4+3
- LiveCodeBench v6Kimi K2 Thinking by 2.583.180.6+2.5
AgentsEven
- GDPValKimi K2 Thinking by 17.924.56.6+17.9
- BrowseCompMiMo V2 Flash by 16.841.558.3+16.8
ReasoningEven
- CritPtMiMo V2 Flash by 1.72.64.3+1.7
- Humanity's Last ExamKimi K2 Thinking by 123.822.8+1
- GPQA Diamondtie83.884.6tie
FactsKimi K2 Thinking
- AA-Omniscience · Non-hallucinationKimi K2 Thinking by 17.824.46.6+17.8
- AA-Omniscience · AccuracyKimi K2 Thinking by 6.130.924.8+6.1
Long documentsEven
- LongBench v2MiMo V2 Flash by 15.545.160.6+15.5
- AA-LCRKimi K2 Thinking by 1.37270.7+1.3
Following instructionsKimi K2 Thinking
- IFBenchKimi K2 Thinking by 3.968.164.2+3.9
Other results20 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- AA Agentic IndexKimi K2 Thinking by 35.947.912+35.9
- xbench-DeepSearchKimi K2 Thinking by 327644+32
- AA-OmniscienceKimi K2 Thinking by 23.9-21.4-45.4+23.9
- Arena-Hard (Hard Prompt)Kimi K2 Thinking by 17.871.954.1+17.8
- Artificial Analysis Coding IndexMiMo V2 Flash by 1534.849.8+15
- BrowseComp-ZHKimi K2 Thinking by 11.162.351.2+11.1
- HMMT Feb. 2025Kimi K2 Thinking by 8.993.384.4+8.9
- Arena-Hard (Creative Writing)MiMo V2 Flash by 6.180.186.2+6.1
- τ²-BenchMiMo V2 Flash by 674.380.3+6
- HMMT 2025Kimi K2 Thinking by 589.484.4+5
- Terminal-Bench 2.0MiMo V2 Flash by 2.835.738.5+2.8
- GAIA (no file)MiMo V2 Flash by 2.675.678.2+2.6
- τ²-Bench Telecom (AA run)MiMo V2 Flash by 29395+2
- browsecomp_with_context_managerKimi K2 Thinking by 1.960.258.3+1.9
- ResearchRubricsKimi K2 Thinking by 1.956.254.3+1.9
- HMMT Nov. 2025MiMo V2 Flash by 1.889.291+1.8
- MRCRMiMo V2 Flash by 1.544.245.7+1.5
- AA IntelligenceKimi K2 Thinking by 1.22220.8+1.2
- AIME 2025tie94.594.1tie
- MMLU-Protie84.684.9tie
Questions people ask
Which is better, Kimi K2 Thinking or MiMo V2 Flash?
Kimi K2 Thinking wins three of the seven areas where both have results: coding, facts and following instructions. MiMo V2 Flash wins math, and costs 87% less. They are level on agents, reasoning and long documents.
Which is better for coding?
Kimi K2 Thinking. It wins 3 of the 5 coding tests both models report; MiMo V2 Flash wins 2.
Which is cheaper?
Kimi K2 Thinking costs $0.60 per million input tokens and $2.50 per million output tokens; MiMo V2 Flash costs $0.10 and $0.30. That makes MiMo V2 Flash about 87% cheaper for the same work.
How do you compare the two?
We use the 36 benchmark tests both models have published scores on. The verdict counts the 16 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 20 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.