Kimi K2.6 vs Mistral Medium 3.5 128B
Wins 7 of 7 areas
Coding · Agents · Reasoning · Facts · Images and charts · Long documents · Following instructions
Wins 0 of 7 areas
—
Kimi K2.6 is the stronger all-rounder.
Scores updated · 25 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.
- CodingWriting and fixing software50Kimi K2.65 of 5 tests
- AgentsCarrying out multi-step tasks on its own30Kimi K2.63 of 4 tests · 1 tie
- ReasoningHard problems that need careful thinking30Kimi K2.63 of 3 tests
- FactsGetting facts right instead of making them up20Kimi K2.62 of 2 tests
- Images and chartsUnderstanding pictures, charts and video20Kimi K2.62 of 2 tests
- Long documentsFinding answers in very long texts10Kimi K2.61 of 1 test
- Following instructionsDoing exactly what it is asked10Kimi K2.61 of 1 test
Long documents 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.
Kimi K2.6 costs 45% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Kimi K2.6 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+41.1points ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+37.7points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+23.7points ahead
Where Mistral Medium 3.5 128B pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 25 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingKimi K2.6
- LMArena · WebDevKimi K2.6 by 244 rating points15091265+244 rating
- Terminal-Bench 2.1Kimi K2.6 by 15.365.950.6+15.3
- SciCodeKimi K2.6 by 11.351.540.2+11.3
- Terminal-Bench HardKimi K2.6 by 10.643.933.3+10.6
- SWE-bench VerifiedKimi K2.6 by 2.680.277.6+2.6
AgentsKimi K2.6
- BrowseCompKimi K2.6 by 37.786.348.6+37.7
- GDPValKimi K2.6 by 13.82713.2+13.8
- τ-Bench V3 · BankingKimi K2.6 by 8.223.315.1+8.2
- Terminal-Bench 4.0tie0.50tie
ReasoningKimi K2.6
- Humanity's Last ExamKimi K2.6 by 23.737.513.8+23.7
- GPQA DiamondKimi K2.6 by 16.391.174.8+16.3
- CritPtKimi K2.6 by 880+8
FactsKimi K2.6
- AA-Omniscience · Non-hallucinationKimi K2.6 by 41.159.518.4+41.1
- AA-Omniscience · AccuracyKimi K2.6 by 7.932.624.7+7.9
Images and chartsKimi K2.6
- MMMU-ProKimi K2.6 by 14.579.464.9+14.5
- LMArena · VisionKimi K2.6 by 60 rating points12821222+60 rating
Other results7 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-OmniscienceKimi K2.6 by 42.15.3-36.8+42.1
- Artificial Analysis Coding IndexKimi K2.6 by 14.961.846.9+14.9
- AA IntelligenceKimi K2.6 by 12.82714.2+12.8
- Finance Agent v2Kimi K2.6 by 12.844.932.1+12.8
- AA Agentic IndexKimi K2.6 by 12.722.19.4+12.7
- τ³-Bench BankingKimi K2.6 by 7.220.613.4+7.2
- τ²-Bench Telecom (AA run)Kimi K2.6 by 1.795.994.2+1.7
Questions people ask
Which is better, Kimi K2.6 or Mistral Medium 3.5 128B?
Kimi K2.6 wins all seven areas where both have results: coding, agents, reasoning, facts, images and charts, long documents and following instructions. Mistral Medium 3.5 128B wins none.
Which is better for coding?
Kimi K2.6. It wins 5 of the 5 coding tests both models report; Mistral Medium 3.5 128B wins none.
Which is cheaper?
Kimi K2.6 costs $0.95 per million input tokens and $4.00 per million output tokens; Mistral Medium 3.5 128B costs $1.50 and $7.50. That makes Kimi K2.6 about 45% cheaper for the same work.
How do you compare the two?
We use the 25 benchmark tests both models have published scores on. The verdict counts the 18 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 7 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.