Big Finance Bench - OpenAI release score
Professional Work
Financial work evaluation in OpenAI's release table without a public task roster or stated scoring recipe. Diagnostic.
Top models (higher is better)
| Model | Score |
|---|---|
| GPT-5.6 Sol | 53.0 |
| GPT-5.6 Terra | 51.0 |
| GPT-5.6 Luna | 36.0 |