GDPVal

GDPVal - agentic benchmark for evaluating general task completion.

Category

General Agents

Max Score

100.0

Score Type

percent

Active

Yes

Model Scores

Scores for GDPVal

Model Score Date Verified
GLM 5.1 Thinking
73.10%
11.06.2026 Unverified
GPT 5.5
100.00%
11.06.2026 Unverified
NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16
26.25%
04.06.2026 Verified
MiniMax-M2.7
28.20%
04.06.2026 Unverified
GLM-5.1
43.60%
04.06.2026 Unverified
Kimi K2.6
34.27%
04.06.2026 Unverified
Qwen3.5-397B-A17B
34.60
04.06.2026 Unverified
DeepSeek-V4-Pro
43.38%
04.06.2026 Unverified
DeepSeek-V4-Flash
33.84%
04.06.2026 Unverified
DeepSeek V4 Pro
77.44%
11.06.2026 Unverified
Claude Sonnet 4.6
89.15%
11.06.2026 Unverified
Claude Opus 4.7
98.05%
11.06.2026 Unverified
MiniMax-M3
87.20%
11.06.2026 Unverified
Gemini 3.1 Pro
50.33%
11.06.2026 Unverified