in the loop

Gemini 3.8 Flash
#20 of 274 scored.

GoogleSep 2, 2026Closed

Capabilities Index

Range 154.4–160.3

156.7

Context

Takes text, image, video, file, audio

1.05M

Input

Per 1M tokens

$0.75

Output

Per 1M tokens

$3.75

Benchmarks

How it scores.

  • GPQA Diamond93.9%

    3rd of 186

  • FrontierMath68.4%

    25th of 81

  • Humanity's Last Exam41.7%

    4th of 40

  • SimpleQA Verified69.7%

    7th of 77

  • Mock AIME98.9%

    16th of 176

Everything else Epoch has for it

  • DTBench92.9%
  • Surface Evolver Bench76.9%
  • DeepSWE73.8%
  • APEX-Agents64.3%
  • LMCA62.2%
  • Chess Puzzles59.0%
  • ProofBench48.0%
  • Mystery Game Puzzles41.6%
  • FrontierCode41.2%
  • FrontierMath-Tier-4-v2-Private21.9%
  • FrontierSWE19.6%
  • Furniture Assembly2.4%

On the index

Around it.

  1. #17GPT-5.4OpenAI · $2.50 in156.8
  2. #18GPT-5.3 CodexOpenAI · $1.75 in156.8
  3. #19Muse Spark 1.3Meta · $1.25 in156.8
  4. #20Gemini 3.8 FlashGoogle · $0.75 in156.7
  5. #21Grok 4.6xAI · $2.00 in156.4
  6. #22Qwen 3.8 MaxAlibaba156.4
  7. #23GPT-5.6 LunaOpenAI · $0.20 in156.4

Sources