# DeepSeek-V3.1 finance AI benchmark results

Every finance benchmark score for DeepSeek-V3.1 by DeepSeek: 4 benchmarks, each with its rank, metric and source date. Scores are never combined across benchmarks.

## Board standings

DeepSeek-V3.1 is first on 0 benchmarks, top 3 on 1 benchmark, scored on 4 benchmarks. Places on each benchmark's main board with 2 or more models; scores are never combined. https://benchmarks.finance/method/#standings
- AI-Trader: #2 of 6

## Scores

| Benchmark | Rank | Score | Metric | Source | Date |
| --- | --- | --- | --- | --- | --- |
| [AI-Trader](https://benchmarks.finance/benchmarks/ai-trader/) | 2 of 6 | 8.4 | cumulative return % | Paper Table 1 | 1 Dec 2025 |
| [StockBench](https://benchmarks.finance/benchmarks/stockbench/) | 8 of 11 | 1.1 | final return % | Paper Table 2 | 2 Mar 2026 |
| [CAIA (Crypto AI Agent Benchmark)](https://benchmarks.finance/benchmarks/caia/) | 8 of 10 | 49.2 | accuracy % (majority vote, 5 runs) | Paper Tables 2-3 | 17 Jan 2026 |
| [Reasoning Models Ace the CFA Exams (CFA mock-exam evaluation)](https://benchmarks.finance/benchmarks/cfa-reasoning-models-ace/) | 6 of 6 | 90.9 | Level I accuracy % | Paper Table 5 | 9 Dec 2025 |

Source: https://benchmarks.finance/models/deepseek-v3-1/

Created and maintained by Moneda.