Gemini 3.6 Flash vs Kimi K3
Wins 2 of 8 areas
Facts · Following instructions
Wins 5 of 8 areas
Coding · Agents · Reasoning · Math · Long documents
Kimi K3 is the stronger all-rounder.Gemini 3.6 Flash is cheaper and better at facts.
Scores updated · 39 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 software05Kimi K35 of 5 tests
- AgentsCarrying out multi-step tasks on its own04Kimi K34 of 4 tests
- ReasoningHard problems that need careful thinking03Kimi K33 of 5 tests · 2 ties
- MathCompetition and research-level math13Kimi K33 of 5 tests · 1 tie
- Long documentsFinding answers in very long texts01Kimi K31 of 1 test
- FactsGetting facts right instead of making them up21Gemini 3.6 Flash2 of 3 tests
- Following instructionsDoing exactly what it is asked10Gemini 3.6 Flash1 of 1 test
- Images and chartsUnderstanding pictures, charts and video11Even1 each
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.
Gemini 3.6 Flash costs 75% 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 K3 pulls ahead
- Fixes real GitHub issues, working as an agentLiveBench · Agentic Coding+18.8points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+17points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+16.1points ahead
Where Gemini 3.6 Flash pulls ahead
- Short factual questions, answered correctlySimpleQA Verified+15.6points ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+4points ahead
- Harder college exam questions with imagesMMMU-Pro+2.7points ahead
Every test, side by side
All 39 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingKimi K3
- LiveBench · Agentic CodingKimi K3 by 18.843.462.2+18.8
- LMArena · WebDevKimi K3 by 122 rating points15361658+122 rating
- Terminal-Bench 2.1Kimi K3 by 7.577.585+7.5
- SciCodeKimi K3 by 6.153.459.5+6.1
- LiveBench · CodingKimi K3 by 3.677.981.5+3.6
AgentsKimi K3
- τ-Bench V3 · BankingKimi K3 by 16.129.946+16.1
- GDPValKimi K3 by 12.539.351.8+12.5
- Terminal-Bench 4.0Kimi K3 by 5.57.112.6+5.5
- OSWorld-VerifiedKimi K3 by 1.88384.8+1.8
ReasoningKimi K3
- CritPtKimi K3 by 12.810.623.4+12.8
- Humanity's Last ExamKimi K3 by 6.140.846.9+6.1
- LiveBench · ReasoningKimi K3 by 5.585.290.7+5.5
- GPQA Diamondtie92.893.5tie
- ARC-AGI-2tie60.460.4tie
FactsGemini 3.6 Flash
- SimpleQA VerifiedGemini 3.6 Flash by 15.666.250.6+15.6
- AA-Omniscience · AccuracyGemini 3.6 Flash by 2.45047.6+2.4
- AA-Omniscience · Non-hallucinationKimi K3 by 2.444.446.8+2.4
Images and chartsEven
- MMMU-ProGemini 3.6 Flash by 2.783.280.5+2.7
- CharXiv (RQ)Kimi K3 by 1.989.491.3+1.9
MathKimi K3
- FrontierMath Tier 4Kimi K3 by 1721.939+17
- FrontierMath Tiers 1-3 (v2)Kimi K3 by 13.25972.2+13.2
- HMMT Feb. 2026Kimi K3 by 7.689.497+7.6
- LiveBench · MathematicsGemini 3.6 Flash by 286.484.4+2
- AIME 2026tie96.796.7tie
Following instructionsGemini 3.6 Flash
- LiveBench · Instruction FollowingGemini 3.6 Flash by 475.471.4+4
Other results13 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.
- GDPval-AA v2 EloKimi K3 by 264 rating points14221686+264 rating
- GDPval-AA v2Kimi K3 by 261 rating points14211682+261 rating
- AA Agentic IndexKimi K3 by 20.430.250.6+20.4
- DeepSWE 1.1Kimi K3 by 204969+20
- DeepSWEKimi K3 by 18.54967.5+18.5
- livebench_data_analysisKimi K3 by 15.76378.7+15.7
- Terminal-bench 3.0Kimi K3 by 125.417.4+12
- AA IntelligenceKimi K3 by 9.63443.6+9.6
- Artificial Analysis Coding IndexKimi K3 by 769.276.2+7
- matharena_visual_math_overallKimi K3 by 3.487.891.3+3.4
- ARC-AGI-1Kimi K3 by 3.391.294.5+3.3
- AA-OmniscienceGemini 3.6 Flash by 2.422.119.7+2.4
- livebench_languageKimi K3 by 1.683.985.5+1.6
Questions people ask
Which is better, Gemini 3.6 Flash or Kimi K3?
Kimi K3 wins five of the eight areas we test: coding, agents, reasoning, math and long documents. Gemini 3.6 Flash wins facts and following instructions, and costs 75% less. They are level on images and charts.
Which is better for coding?
Kimi K3. It wins 5 of the 5 coding tests both models report; Gemini 3.6 Flash wins none.
Which is cheaper?
Gemini 3.6 Flash costs $0.75 per million input tokens and $3.75 per million output tokens; Kimi K3 costs $3.00 and $15.00. That makes Gemini 3.6 Flash about 75% cheaper for the same work.
How do you compare the two?
We use the 39 benchmark tests both models have published scores on. The verdict counts the 26 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 13 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.