Kimi K2.5 vs o4-mini
Wins 6 of 7 areas
Coding · Reasoning · Facts · Images and charts · Long documents · Following instructions
Wins 0 of 7 areas
—
Kimi K2.5 is the stronger all-rounder.
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
- FactsGetting facts right instead of making them up40Kimi K2.54 of 4 tests
- Images and chartsUnderstanding pictures, charts and video40Kimi K2.54 of 4 tests
- CodingWriting and fixing software30Kimi K2.53 of 3 tests
- Following instructionsDoing exactly what it is asked20Kimi K2.52 of 2 tests
- Long documentsFinding answers in very long texts10Kimi K2.51 of 1 test
- AgentsCarrying out multi-step tasks on its own11Even1 each
Long documents rests on a single test.
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 35% 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 K2.5 pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+23.4points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+19.6points ahead
- Keeps track of context across a multi-turn chatMulti-Challenge+18.4points ahead
Where o4-mini pulls ahead
- Real work tasks from 44 professionsGDPVal+8.2points 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
- Terminal-Bench HardKimi K2.5 by 19.634.815.2+19.6
- SWE-bench VerifiedKimi K2.5 by 8.776.868.1+8.7
- SciCodeKimi K2.5 by 2.54946.5+2.5
AgentsEven
- BrowseCompKimi K2.5 by 23.474.951.5+23.4
- GDPValo4-mini by 8.217.225.4+8.2
ReasoningKimi K2.5
- Humanity's Last ExamKimi K2.5 by 14.230.716.5+14.2
- GPQA DiamondKimi K2.5 by 9.587.978.4+9.5
- ARC-AGI-2Kimi K2.5 by 5.711.86.1+5.7
- CritPtKimi K2.5 by 2.53.10.6+2.5
FactsKimi K2.5
- SimpleQA VerifiedKimi K2.5 by 15.534.318.8+15.5
- AA-Omniscience · Non-hallucinationKimi K2.5 by 14.834.319.5+14.8
- AA-Omniscience · AccuracyKimi K2.5 by 10.435.224.8+10.4
- Vectara HHEM hallucination ratelower is betterKimi K2.5 by 4.414.218.6+4.4
Images and chartsKimi K2.5
- LMArena · VisionKimi K2.5 by 75 rating points12691194+75 rating
- MMMU-ProKimi K2.5 by 6.275.469.2+6.2
- MathVistaKimi K2.5 by 5.890.184.3+5.8
- CharXiv (RQ)Kimi K2.5 by 5.577.572+5.5
Following instructionsKimi K2.5
- Multi-ChallengeKimi K2.5 by 18.461.443+18.4
- IFBenchKimi K2.5 by 1.570.268.7+1.5
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.
- τ²-Bench Telecom (AA run)Kimi K2.5 by 40.395.955.6+40.3
- AA-OmniscienceKimi K2.5 by 28.4-7.3-35.7+28.4
- Artificial Analysis Coding IndexKimi K2.5 by 21.246.825.6+21.2
- AA Agentic Indexo4-mini by 14.421.736.1+14.4
- vectara_answer_rateo4-mini by 792.299.2+7
- AA IntelligenceKimi K2.5 by 6.823.516.7+6.8
- ARC-AGI-1Kimi K2.5 by 6.665.358.7+6.6
- vectara_factual_consistencyKimi K2.5 by 4.485.881.4+4.4
- StrongREJECTKimi K2.5 by 3.599.596+3.5
- AIME 2025Kimi K2.5 by 3.496.192.7+3.4
- frontiermath_tier_4_v1o4-mini by 2.14.26.3+2.1
- vectara_avg_summary_lengtho4-mini by 15.7 rating points112127.7+15.7 rating
Questions people ask
Which is better, Kimi K2.5 or o4-mini?
Kimi K2.5 wins six of the seven areas where both have results: coding, reasoning, facts, images and charts, long documents and following instructions. o4-mini wins none. They are level on agents.
Which is better for coding?
Kimi K2.5. It wins 3 of the 3 coding tests both models report; o4-mini wins none.
Which is cheaper?
Kimi K2.5 costs $0.60 per million input tokens and $3.00 per million output tokens; o4-mini costs $1.10 and $4.40. That makes Kimi K2.5 about 35% 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.