Gemini 3 Pro vs Kimi K2 Thinking
Wins 6 of 7 areas
Coding · Agents · Reasoning · Math · Long documents · Following instructions
Wins 0 of 7 areas
—
Gemini 3 Pro is the stronger all-rounder.Kimi K2 Thinking is cheaper.
Scores updated · 38 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 software50Gemini 3 Pro5 of 5 tests
- ReasoningHard problems that need careful thinking40Gemini 3 Pro4 of 4 tests
- AgentsCarrying out multi-step tasks on its own20Gemini 3 Pro2 of 2 tests
- Long documentsFinding answers in very long texts20Gemini 3 Pro2 of 2 tests
- Following instructionsDoing exactly what it is asked20Gemini 3 Pro2 of 2 tests
- MathCompetition and research-level math10Gemini 3 Pro1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
Math 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 Thinking costs 78% 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 Gemini 3 Pro pulls ahead
- Common-sense trick questionsSimpleBench+50.1points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+24.9points ahead
- Questions about very long textsLongBench v2+23.1points ahead
Where Kimi K2 Thinking pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+15.9points ahead
Every test, side by side
All 38 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemini 3 Pro
- SciCodeGemini 3 Pro by 13.756.142.4+13.7
- SWE-bench VerifiedGemini 3 Pro by 12.876.263.4+12.8
- Terminal-Bench HardGemini 3 Pro by 10.641.731.1+10.6
- LiveCodeBench v6Gemini 3 Pro by 4.387.483.1+4.3
- SWE-bench MultilingualGemini 3 Pro by 3.96561.1+3.9
AgentsGemini 3 Pro
- BrowseCompGemini 3 Pro by 17.759.241.5+17.7
- GDPValGemini 3 Pro by 9.734.224.5+9.7
ReasoningGemini 3 Pro
- SimpleBenchGemini 3 Pro by 50.176.426.3+50.1
- Humanity's Last ExamGemini 3 Pro by 15.939.723.8+15.9
- GPQA DiamondGemini 3 Pro by 790.883.8+7
- CritPtGemini 3 Pro by 6.59.12.6+6.5
FactsEven
- AA-Omniscience · AccuracyGemini 3 Pro by 24.955.830.9+24.9
- AA-Omniscience · Non-hallucinationKimi K2 Thinking by 15.98.524.4+15.9
Long documentsGemini 3 Pro
- LongBench v2Gemini 3 Pro by 23.168.245.1+23.1
- AA-LCRGemini 3 Pro by 47672+4
Following instructionsGemini 3 Pro
- Multi-ChallengeGemini 3 Pro by 10.365.755.4+10.3
- IFBenchGemini 3 Pro by 2.370.468.1+2.3
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.
- MRCRGemini 3 Pro by 45.589.744.2+45.5
- AA-OmniscienceGemini 3 Pro by 36.715.3-21.4+36.7
- τ²-BenchGemini 3 Pro by 23.79874.3+23.7
- OJ-Bench (cpp)Gemini 3 Pro by 19.868.548.7+19.8
- frontiermath_tier_4_v1Gemini 3 Pro by 18.818.80+18.8
- Terminal-Bench 2.0Gemini 3 Pro by 18.554.235.7+18.5
- Arena-Hard (Creative Writing)Gemini 3 Pro by 13.593.680.1+13.5
- LiveCodeBenchGemini 3 Pro by 12.89279.2+12.8
- Artificial Analysis Coding IndexGemini 3 Pro by 11.746.534.8+11.7
- AA IntelligenceGemini 3 Pro by 62822+6
- τ²-Bench Telecom (AA run)Kimi K2 Thinking by 5.987.193+5.9
- AIME 2025Gemini 3 Pro by 5.510094.5+5.5
- MMLU-ProGemini 3 Pro by 5.590.184.6+5.5
- HMMT Feb. 2025Gemini 3 Pro by 4.297.593.3+4.2
- AA Agentic IndexGemini 3 Pro by 4.15247.9+4.1
- HMMT Nov. 2025Gemini 3 Pro by 4.193.389.2+4.1
- AIME25 no toolsGemini 3 Pro by 1.59694.5+1.5
- browsecomp_with_context_managerKimi K2 Thinking by 159.260.2+1
- HLE (with tools)tie45.844.9tie
- Arena-Hard (Hard Prompt)tie72.671.9tie
Questions people ask
Which is better, Gemini 3 Pro or Kimi K2 Thinking?
Gemini 3 Pro wins six of the seven areas where both have results: coding, agents, reasoning, math, long documents and following instructions. Kimi K2 Thinking wins none, but costs 78% less. They are level on facts.
Which is better for coding?
Gemini 3 Pro. It wins 5 of the 5 coding tests both models report; Kimi K2 Thinking wins none.
Which is cheaper?
Gemini 3 Pro costs $2.00 per million input tokens and $12.00 per million output tokens; Kimi K2 Thinking costs $0.60 and $2.50. That makes Kimi K2 Thinking about 78% cheaper for the same work.
How do you compare the two?
We use the 38 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 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.