Inference

Agents' Last Exam

An agent benchmark reported in DeepSeek's V4 Flash 0731 launch comparison.

  • In index
  • Agentic
ModelsModels with a result.15
Top resultBest result on this test.59.3%GPT-6 Astra
Top-3 spreadPoints from first to third.6.9 pts
YearYear of release.N/A

Result and price

020406080100$0.1$0.3$1$3$10$30$100RESULTOUTPUT PRICE PER 1M TOKENS · LOG SCALE

Results

15 results
#ModelResultThe score from the source.
01GPT-6 Astra59.3%
02GPT-6 Sol56.4%
03Qwen3.8 Max52.4%
04Qwen3.8-Flash-Next51.2%
05Qwen3.8-27B42.9%
06DeepSeek V4.1 Flash31.8%
07MiMo-V2.6-Pro31.6%
08Step 5 Preview29.5%
09GLM-5.328.5%
10MiMo-V2.6-Flash27.6%
11Gemini 3.7 Flash26.3%
11GLM-5.3-Flash26.3%
13DeepSeek V4 Pro 081325.7%
14DeepSeek V4 Flash 073125.2%
15Hy4 preview22.8%