in the loop

Kimi K2.6
#50 of 274 scored.

MoonshotApr 20, 2026Open weights

Capabilities Index

Range 149.3–152.8

151.1

Context

Takes text, image

262K

Input

Per 1M tokens

$0.47

Output

Per 1M tokens

$2.45

Benchmarks

How it scores.

  • GPQA Diamond87.7%

    34th of 186

  • SWE-bench Verified76.6%

    9th of 32

  • FrontierMath57.2%

    38th of 81

  • SimpleQA Verified34.9%

    50th of 77

  • Mock AIME96.1%

    29th of 176

Everything else Epoch has for it

  • DTBench84.9%
  • WeirdML55.9%
  • LMCA43.8%
  • FrontierMath-Tier-4-v2-Private25.6%
  • Chess Puzzles22.1%
  • ExploitBench18.4%
  • ProofBench16.0%
  • Mystery Game Puzzles9.7%
  • OSWorld 2.04.6%
  • EBR-bench2.4%
  • Furniture Assembly0.0%

On the index

Around it.

  1. #47GLM-5.3-FlashZ.ai · $0.15 in151.9
  2. #48Gemini 3 FlashGoogle151.8
  3. #49GLM-5.2Z.ai · $0.06 in151.8
  4. #50Kimi K2.6Moonshot · $0.47 in151.1
  5. #51GPT-5 ProOpenAI · $15 in150.3
  6. #52Inkling-SmallThinking Machines · $0.45 in150.2
  7. #53Claude Opus 4.5Anthropic · $5.00 in150.1

Sources