DeepSearch QA
DeepSearch QA - full-task agentic benchmark measuring ability to work within scaffolds, write and debug code, and resolve multi-turn requests from start to finish.
Category
General Agents
Max Score
100.0
Score Type
percent
Active
Yes
Model Scores
Scores for DeepSearch QA
| Model | Score | Date | Verified |
|---|---|---|---|
| Gemma4-31B-it |
61.70
|
01.08.2026 | Unverified |
| Qwen3.6-27B |
28.23%
|
01.08.2026 | Unverified |
| Kimi K3 |
100.00%
|
18.08.2026 | Verified |
| GPT-5.4 |
50.75%
|
23.08.2026 | Verified |
| Claude Opus 4.6 |
88.89%
|
23.08.2026 | Verified |
| Gemini 3.1 Pro |
60.66%
|
23.08.2026 | Verified |
| Kimi K2.5 |
81.98%
|
23.08.2026 | Verified |
| Muse-Glimmer-30B |
38.74%
|
01.08.2026 | Verified |
| Kimi K2.6 |
92.49%
|
23.08.2026 | Verified |