Comparing 2 models

Gemini 3.1 Pro Preview vs Kimi K3

Benchmarks through

Verdict

Gemini 3.1 Pro Preview places higher on 3 of the 3 tasks where both are ranked. Kimi K3 costs 10.4x less per input token. Only Kimi K3 has open weights for self-hosting.

  • Choose Gemini 3.1 Pro Preview for Translation (#5), Literary Translation (#5), Everyday Translation (#5).
  • Choose Kimi K3 for you need to self-host.

Built from the published task rankings below; ranks come from each task's own weighted benchmarks, so a win on one task says nothing about another.

metric
Model 1Gemini 3.1 Pro Preview
Model 2Kimi K3
Best for
Business Writing (#3)
JSON & Schema Output (#2)
Provider
google
moonshotai
Context
1.0MBest
1.0MBest
Input / 1M
$2
$0.19Best
Output / 1M
$12Best
$16
Tool use
yesBest
yesBest
Vision
yesBest
yesBest
Reasoning
yesBest
yesBest
Open-weight
—
yesBest
Released
Feb 19, 2026
Jul 16, 2026

Measured head to head

Benchmarks both models were run on

EvaluationGemini 3.1 Pro PreviewKimi K3Source
Design Arena Coding1220 Elo1352 Elosource · 2026-11-15
Design Arena Full-stack Agent1041 Elo1289 Elosource · 2026-11-15
Design Arena UI1215 Elo1342 Elosource · 2026-11-15
Design Arena Web-app Agent1103 Elo1273 Elosource · 2026-11-15
Design Arena Website1225 Elo1336 Elosource · 2026-11-15
FrontierMath Tiers 1-3 (Epoch AI)59.6% ± 2.972.2% ± 2.7source · 2026-06-11
GPQA Diamond (Epoch AI)94.4% ± 1.693.1% ± 1.5source · 2026-08-06
GPQA Diamond (Epoch AI)94.1% ± 1.793.1% ± 1.5source · 2026-02-20
LiveBench Agentic Coding44.1%62.2%source · 2026-10-18
LiveBench Coding76.5%81.5%source · 2026-10-18
LiveBench Data Analysis78.5%78.7%source · 2026-10-18
LiveBench Instruction Following79.1%71.4%source · 2026-10-18
LiveBench Language85.4%85.5%source · 2026-10-18
LiveBench Math91.0%84.4%source · 2026-10-18
LiveBench Reasoning84%90.7%source · 2026-10-18
LMArena Agent-4.50.6source · 2026-11-11
LMArena Agent Bash Recovery-23.34.4source · 2026-10-08
LMArena Agent Steerability-8.2-0.8source · 2026-11-11
LMArena Agent Task Outcome-11.33.4source · 2026-11-11
LMArena Agent Tool Hallucination-0.41source · 2026-11-11
LMArena Creative Writing1483 Elo1453 Elosource · 2026-11-11
LMArena Instruction Following1465 Elo1484 Elosource · 2026-11-11
LMArena Long Query1482 Elo1490 Elosource · 2026-11-11
LMArena Math1485 Elo1496 Elosource · 2026-11-11
LMArena Text1487 Elo1487 Elosource · 2026-11-11
LMArena Text Factuality1472 Elo1471 Elosource · 2026-11-11
LMArena WebDev1446 Elo1658 Elosource · 2026-11-11
LMArena WebDev Image-to-Web1480 Elo1586 Elosource · 2026-11-10
OTIS Mock AIME (Epoch AI)95.6% ± 3.197.2% ± 1.1source · 2026-08-06
OTIS Mock AIME (Epoch AI)95.6% ± 3.197.2% ± 1.1source · 2026-02-20
SimpleQA Verified (Epoch AI)73.5% ± 1.450.6% ± 1.6source · 2026-08-10
tau2-bench (1.0.1)26%37.1%source · 2026-05-05

Values are as published by each evaluator, in its own units; rows from different evaluators are not on a common scale.