in the loop

Muse Spark
#45 of 274 scored.

MetaApr 8, 2026Closed

Capabilities Index

Range 150.1–155.5

152.0

Context

Tokens

—

Input

Per 1M tokens

—

Output

Per 1M tokens

—

Benchmarks

How it scores.

  • GPQA Diamond86.4%

    42nd of 186

  • Humanity's Last Exam37.6%

    6th of 40

  • Mock AIME88.9%

    53rd of 176

Everything else Epoch has for it

  • FrontierMath-2025-02-28-Private68.4%
  • FrontierMath-Tier-4-2025-07-01-Private24.3%
  • ProofBench17.0%

On the index

Around it.

  1. #42GPT-5.2OpenAI · $1.75 in153.4
  2. #43Gemini 3 ProGoogle152.9
  3. #44Claude Sonnet 4.6Anthropic · $3.00 in152.2
  4. #45Muse SparkMeta152.0
  5. #46Grok 4.20xAI · $1.25 in152.0
  6. #47GLM-5.3-FlashZ.ai · $0.15 in151.9
  7. #48Gemini 3 FlashGoogle151.8

Sources