DeepSeek V4 Pro (0813) (Non-reasoning): benchmark results
Provider: DeepSeek. Released 2026-08-13. Access: Open.
Aggregate Rating 1647 ± 14, rank #554 of 2245 rated models, from 84 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA-Omniscience Accuracy | 41.45 | Accuracy (%) | 78.7 |
| German LLM Index - INCLUDE v1.2 | 87.85 | Exact match, strict final-answer parsing (%) | 75 |
| German LLM Index - ScaLA Acceptability | 75.61 | Matthews correlation coefficient (x100) | 67.4 |
| German LLM Index - MMLU-ProX v1.1 | 80.73 | Accuracy (%) | 66.7 |
| AA Omniscience | -12.38 | Score | 61.3 |
| AA-Briefcase - PDF - Rubric Pass Rate | 35.34 | Rubric Pass Rate (%) | 53.2 |
| AutomationBench-AA - Strict | 14.61 | Strict Task Success (%) | 52.9 |
| AA-Briefcase - DOCX - Rubric Pass Rate | 34.32 | Rubric Pass Rate (%) | 51.9 |
| AA Humanity's Last Exam | 10.57 | Accuracy (%) | 49.3 |
| AA-Omniscience v1.1 Accuracy | 38.6 | Accuracy (%) | 46.5 |
| AA GDPval | 1065.08 | ELO | 45.9 |
| AutomationBench-AA | 34.91 | Guardrail-Adjusted Objective Completion (%) | 45.4 |
Interactive version: theaggregate.ai/model?slug=deepseek-v4-pro-0813-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-16.