Kimi K2.5 vs o3
Wins 5 of 7 areas
Coding · Agents · Reasoning · Images and charts · Long documents
Wins 1 of 7 areas
Facts
Kimi K2.5 is the stronger all-rounder.o3 is better at facts.
Scores updated · 32 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.
- ReasoningHard problems that need careful thinking40Kimi K2.54 of 4 tests
- Images and chartsUnderstanding pictures, charts and video31Kimi K2.53 of 4 tests
- AgentsCarrying out multi-step tasks on its own20Kimi K2.52 of 2 tests
- Long documentsFinding answers in very long texts20Kimi K2.52 of 2 tests
- CodingWriting and fixing software21Kimi K2.52 of 3 tests
- FactsGetting facts right instead of making them up12o32 of 3 tests
- Following instructionsDoing exactly what it is asked11Even1 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 K2.5 costs 64% 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 K2.5 pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+25.2points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+22.4points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+10.6points ahead
Where o3 pulls ahead
- Short factual questions, answered correctlySimpleQA Verified+15.1points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+3.4points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+2.3points ahead
Every test, side by side
All 32 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingKimi K2.5
- SciCodeKimi K2.5 by 84941+8
- SWE-bench VerifiedKimi K2.5 by 7.776.869.1+7.7
- Terminal-Bench Hardo3 by 2.334.837.1+2.3
AgentsKimi K2.5
- BrowseCompKimi K2.5 by 25.274.949.7+25.2
- GDPValKimi K2.5 by 4.417.212.8+4.4
ReasoningKimi K2.5
- Humanity's Last ExamKimi K2.5 by 10.630.720.1+10.6
- ARC-AGI-2Kimi K2.5 by 5.311.86.5+5.3
- GPQA DiamondKimi K2.5 by 5.287.982.7+5.2
- CritPtKimi K2.5 by 23.11.1+2
Factso3
- AA-Omniscience · Non-hallucinationKimi K2.5 by 22.434.311.9+22.4
- SimpleQA Verifiedo3 by 15.134.349.4+15.1
- AA-Omniscience · Accuracyo3 by 3.435.238.6+3.4
Images and chartsKimi K2.5
- LMArena · VisionKimi K2.5 by 55 rating points12691214+55 rating
- MMMU-ProKimi K2.5 by 5.375.470.1+5.3
- MathVistaKimi K2.5 by 3.390.186.8+3.3
- CharXiv (RQ)o3 by 1.177.578.6+1.1
Long documentsKimi K2.5
- AA-LCRKimi K2.5 by 3.37874.7+3.3
- LongBench v2Kimi K2.5 by 2.26158.8+2.2
Following instructionsEven
- IFBencho3 by 1.270.271.4+1.2
- Multi-ChallengeKimi K2.5 by 161.460.4+1
Other results12 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.
- HMMT Feb. 2025Kimi K2.5 by 17.995.477.5+17.9
- τ²-Bench Telecom (AA run)Kimi K2.5 by 15.295.980.7+15.2
- AA Agentic Indexo3 by 14.421.736.1+14.4
- AIME 2025Kimi K2.5 by 9.796.186.4+9.7
- Artificial Analysis Coding IndexKimi K2.5 by 8.446.838.4+8.4
- AA-OmniscienceKimi K2.5 by 8.2-7.3-15.6+8.2
- ARC-AGI-1Kimi K2.5 by 4.565.360.8+4.5
- AA IntelligenceKimi K2.5 by 3.323.520.2+3.3
- VideoMMMUKimi K2.5 by 3.386.683.3+3.3
- StrongREJECTKimi K2.5 by 2.599.597+2.5
- frontiermath_tier_4_v1Kimi K2.5 by 2.14.22.1+2.1
- MMLU-ProKimi K2.5 by 2.187.185+2.1
Questions people ask
Which is better, Kimi K2.5 or o3?
Kimi K2.5 wins five of the seven areas where both have results: coding, agents, reasoning, images and charts and long documents. o3 wins facts. They are level on following instructions.
Which is better for coding?
Kimi K2.5. It wins 2 of the 3 coding tests both models report; o3 wins 1.
Which is cheaper?
Kimi K2.5 costs $0.60 per million input tokens and $3.00 per million output tokens; o3 costs $2.00 and $8.00. That makes Kimi K2.5 about 64% cheaper for the same work.
How do you compare the two?
We use the 32 benchmark tests both models have published scores on. The verdict counts the 20 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 12 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.