Kimi K2.6 vs o3
Wins 7 of 8 areas
Coding · Agents · Reasoning · Images and charts · Math · Long documents · Following instructions
Wins 1 of 8 areas
Facts
Kimi K2.6 is the stronger all-rounder.o3 is better at facts.
Scores updated · 26 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. 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 software30Kimi K2.63 of 3 tests
- ReasoningHard problems that need careful thinking30Kimi K2.63 of 3 tests
- Images and chartsUnderstanding pictures, charts and video30Kimi K2.63 of 3 tests
- AgentsCarrying out multi-step tasks on its own20Kimi K2.62 of 2 tests
- MathCompetition and research-level math10Kimi K2.61 of 1 test
- Long documentsFinding answers in very long texts10Kimi K2.61 of 1 test
- Following instructionsDoing exactly what it is asked10Kimi K2.61 of 1 test
- FactsGetting facts right instead of making them up12o32 of 3 tests
Math, 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 51% 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+47.6points ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+36.6points ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+23.9points ahead
Where o3 pulls ahead
- Short factual questions, answered correctlySimpleQA Verified+14.5points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+6points ahead
Every test, side by side
All 26 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingKimi K2.6
- SWE-bench VerifiedKimi K2.6 by 11.180.269.1+11.1
- SciCodeKimi K2.6 by 10.551.541+10.5
- Terminal-Bench HardKimi K2.6 by 6.843.937.1+6.8
AgentsKimi K2.6
- BrowseCompKimi K2.6 by 36.686.349.7+36.6
- GDPValKimi K2.6 by 14.22712.8+14.2
ReasoningKimi K2.6
- Humanity's Last ExamKimi K2.6 by 17.437.520.1+17.4
- GPQA DiamondKimi K2.6 by 8.491.182.7+8.4
- CritPtKimi K2.6 by 6.981.1+6.9
Factso3
- AA-Omniscience · Non-hallucinationKimi K2.6 by 47.659.511.9+47.6
- SimpleQA Verifiedo3 by 14.534.949.4+14.5
- AA-Omniscience · Accuracyo3 by 632.638.6+6
Images and chartsKimi K2.6
- MMMU-ProKimi K2.6 by 9.379.470.1+9.3
- CharXiv (RQ)Kimi K2.6 by 8.186.778.6+8.1
- LMArena · VisionKimi K2.6 by 68 rating points12821214+68 rating
Other results9 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.
- Artificial Analysis Coding IndexKimi K2.6 by 23.461.838.4+23.4
- AA-OmniscienceKimi K2.6 by 20.85.3-15.6+20.8
- τ²-Bench Telecom (AA run)Kimi K2.6 by 15.295.980.7+15.2
- AA Agentic Indexo3 by 1422.136.1+14
- frontiermath_tier_4_v1Kimi K2.6 by 12.514.62.1+12.5
- AA IntelligenceKimi K2.6 by 6.82720.2+6.8
- LiveCodeBenchKimi K2.6 by 4.989.684.7+4.9
- StrongREJECTKimi K2.6 by 2.899.897+2.8
- MMLU-ProKimi K2.6 by 2.187.185+2.1
Questions people ask
Which is better, Kimi K2.6 or o3?
Kimi K2.6 wins seven of the eight areas we test: coding, agents, reasoning, images and charts, math, long documents and following instructions. o3 wins facts.
Which is better for coding?
Kimi K2.6. It wins 3 of the 3 coding tests both models report; o3 wins none.
Which is cheaper?
Kimi K2.6 costs $0.95 per million input tokens and $4.00 per million output tokens; o3 costs $2.00 and $8.00. That makes Kimi K2.6 about 51% cheaper for the same work.
How do you compare the two?
We use the 26 benchmark tests both models have published scores on. The verdict counts the 17 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 9 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.