Benchmark
finsearchcomp-global
Agentic benchmarks measure how well models carry out tasks that take several steps. Scores are percentages.
All agentic benchmarksScores from FinSearchComp docs.
- Models measured
- 6
- Scores recorded
- 6
- Sources
- 1
- IQ weight
- 0.48
Score distribution
- 43.3%
- Mean
- 42.6%
- Median
- 18.2%
- Std dev
- 29.5%
- 10th percentile
- 68.9%
- 90th percentile
- 17.2%
- Min
This benchmark has a weight of 0.48 for chat models, measured over 6 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Grok 4· FinSearchComp docs | 68.9% |
| 2 | GPT 5thinking· FinSearchComp docs | 63.9% |
| 3 | Gemini 2.5 Pro· FinSearchComp docs | 42.6% |
| 4 | Qwen3 235B A22B Instruct 2507· FinSearchComp docs | 37.4% |
| 5 | Kimi K2· FinSearchComp docs | 29.5% |
| 6 | DeepSeek R1· FinSearchComp docs | 17.2% |
Data credits: FinSearchComp, CC BY 4.0 (licence)