# GPT-4o vs Llama-3.3-70B

Every finance benchmark where both have a score (OpenAI and Meta). Data refreshed 2 Nov 2026. Scores from different benchmarks use different scales and are not added up.

| Benchmark | Metric | GPT-4o | GPT-4o source | Llama-3.3-70B | Llama-3.3-70B source | Ahead |
| --- | --- | --- | --- | --- | --- | --- |
| FinLFQA | evidence attribution F1 (higher is better) | 56.5 | Paper Table 6, 7 Oct 2025, https://arxiv.org/html/2510.06426v1 | 52.9 | Paper Table 6, 7 Oct 2025, https://arxiv.org/html/2510.06426v1 | GPT-4o |
| FinLFQA | LLM-as-judge answer score (/15) (higher is better) | 13.5 | Paper Table 6, 7 Oct 2025, https://arxiv.org/html/2510.06426v1 | 12.9 | Paper Table 6, 7 Oct 2025, https://arxiv.org/html/2510.06426v1 | GPT-4o |
| K-FinHallu | binary detection F1 (higher is better) | 0.733 | Paper Table 3, 28 May 2026, https://arxiv.org/html/2605.29523v1 | 0.756 | Paper Table 3, 28 May 2026, https://arxiv.org/html/2605.29523v1 | Llama-3.3-70B |

Source: https://benchmarks.finance/compare/gpt-4o-vs-llama-3-3-70b/

Created and maintained by Moneda.