in the loop

DeepSeek-V3.1
#119 of 274 scored.

DeepSeekAug 21, 2025Open weights

Capabilities Index

Range 135.4–143.2

139.9

Context

Takes text

164K

Input

Per 1M tokens

$0.25

Output

Per 1M tokens

$0.95

Benchmarks

No independent scores yet.

Epoch AI hasn't run DeepSeek-V3.1 yet. Its scores show up here the morning after they do.

Everything else Epoch has for it

  • DTBench71.1%
  • Fiction.LiveBench52.8%
  • WeirdML38.4%
  • LMCA28.6%
  • SimpleBench28.0%

On the index

Around it.

  1. #116Kimi K2 (Jul 2025)Moonshot140.1
  2. #117Gemini 2.5 Flash (Apr 2025)Google140.0
  3. #118gpt-oss-120bOpenAI · $0.037 in139.9
  4. #119DeepSeek-V3.1DeepSeek · $0.25 in139.9
  5. #120Qwen3-30B-A3B-Thinking (Jul 2025)Alibaba139.6
  6. #121Qwen3.5-9BAlibaba · $0.10 in139.5
  7. #122GPT-5 nanoOpenAI · $0.05 in139.4

Sources