Gemini 3 Pro vs Kimi K2 Instruct
Wins 7 of 7 areas
Coding · Agents · Reasoning · Facts · Math · Long documents · Following instructions
Wins 0 of 7 areas
—
Gemini 3 Pro is the stronger all-rounder.Kimi K2 Instruct is cheaper.
Scores updated · 42 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 software60Gemini 3 Pro6 of 6 tests
- ReasoningHard problems that need careful thinking30Gemini 3 Pro3 of 3 tests
- AgentsCarrying out multi-step tasks on its own20Gemini 3 Pro2 of 2 tests
- Following instructionsDoing exactly what it is asked20Gemini 3 Pro2 of 2 tests
- FactsGetting facts right instead of making them up21Gemini 3 Pro2 of 3 tests
- MathCompetition and research-level math10Gemini 3 Pro1 of 1 test
- Long documentsFinding answers in very long texts10Gemini 3 Pro1 of 1 test
Math and long documents 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 Instruct 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
- Finds hard-to-locate facts by browsing the webBrowseComp+45.1points ahead
- Olympiad math problems with short answersIMOAnswerBench+37.3points ahead
- Recent programming contest problemsLiveCodeBench v6+33.7points ahead
Where Kimi K2 Instruct pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+21.8points ahead
Every test, side by side
All 42 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemini 3 Pro
- LiveCodeBench v6Gemini 3 Pro by 33.787.453.7+33.7
- SciCodeGemini 3 Pro by 25.456.130.7+25.4
- Terminal-Bench HardGemini 3 Pro by 18.241.723.5+18.2
- SWE-bench MultilingualGemini 3 Pro by 17.76547.3+17.7
- SWE-bench ProGemini 3 Pro by 15.643.327.7+15.6
- SWE-bench VerifiedGemini 3 Pro by 10.476.265.8+10.4
AgentsGemini 3 Pro
- BrowseCompGemini 3 Pro by 45.159.214.1+45.1
- GDPValGemini 3 Pro by 1634.218.2+16
ReasoningGemini 3 Pro
- Humanity's Last ExamGemini 3 Pro by 33.339.76.4+33.3
- GPQA DiamondGemini 3 Pro by 14.190.876.7+14.1
- CritPtGemini 3 Pro by 9.19.10+9.1
FactsGemini 3 Pro
- AA-Omniscience · AccuracyGemini 3 Pro by 30.455.825.4+30.4
- AA-Omniscience · Non-hallucinationKimi K2 Instruct by 21.88.530.3+21.8
- Vectara HHEM hallucination ratelower is betterGemini 3 Pro by 4.313.617.9+4.3
Following instructionsGemini 3 Pro
- IFBenchGemini 3 Pro by 28.770.441.7+28.7
- Multi-ChallengeGemini 3 Pro by 11.665.754.1+11.6
Other results24 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.
- AIME 2025Gemini 3 Pro by 50.510049.5+50.5
- AIME25 no toolsGemini 3 Pro by 459651+45
- OJ-Bench (cpp)Gemini 3 Pro by 4368.525.5+43
- AA-OmniscienceGemini 3 Pro by 41.915.3-26.6+41.9
- SimpleQAGemini 3 Pro by 41.172.131+41.1
- LiveCodeBenchGemini 3 Pro by 38.39253.7+38.3
- τ²-BenchGemini 3 Pro by 32.29865.8+32.2
- HarmBenchKimi K2 Instruct by 24.972.597.4+24.9
- Artificial Analysis Coding IndexGemini 3 Pro by 20.646.525.9+20.6
- HLE (with tools)Gemini 3 Pro by 18.945.826.9+18.9
- τ²-Bench (Retail)Gemini 3 Pro by 14.785.370.6+14.7
- AA Agentic IndexGemini 3 Pro by 14.35237.7+14.3
- τ²-Bench Telecom (AA run)Gemini 3 Pro by 13.787.173.4+13.7
- AA IntelligenceGemini 3 Pro by 12.72815.3+12.7
- Multi-SWE-BenchGemini 3 Pro by 9.242.733.5+9.2
- MMLU-ProGemini 3 Pro by 990.181.1+9
- vectara_factual_consistencyGemini 3 Pro by 4.386.482.1+4.3
- vectara_avg_summary_lengthGemini 3 Pro by 42.7 rating points101.959.2+42.7 rating
- LiveBenchKimi K2 Instruct by 373.476.4+3
- simple_safety_testsKimi K2 Instruct by 2.597.5100+2.5
- anthropic_red_teamKimi K2 Instruct by 2.29799.3+2.2
- AIR-Bench 2024tie73.274.1tie
- XSTesttie97.398.2tie
- vectara_answer_ratetie99.498.6tie
Questions people ask
Which is better, Gemini 3 Pro or Kimi K2 Instruct?
Gemini 3 Pro wins all seven areas where both have results: coding, agents, reasoning, facts, math, long documents and following instructions. Kimi K2 Instruct wins none, but costs 78% less.
Which is better for coding?
Gemini 3 Pro. It wins 6 of the 6 coding tests both models report; Kimi K2 Instruct wins none.
Which is cheaper?
Gemini 3 Pro costs $2.00 per million input tokens and $12.00 per million output tokens; Kimi K2 Instruct costs $0.60 and $2.50. That makes Kimi K2 Instruct about 78% cheaper for the same work.
How do you compare the two?
We use the 42 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 24 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.