DeepSeek V4.1 Flash (1017) (Non-reasoning): benchmark results
Provider: DeepSeek. Released 2026-10-17. Access: Open.
Aggregate Rating 1708 ± 16, rank #336 of 2170 rated models, from 71 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA-Omniscience Hallucination Rate | 25.3 | Hallucination Rate (%) | 86.2 |
| AA-Briefcase - Presentation Elo | 1459.8 | Elo | 84.7 |
| AA-Briefcase - DOCX - Rubric Pass Rate | 47.93 | Rubric Pass Rate (%) | 78.4 |
| AA-Briefcase - PPTX - Rubric Pass Rate | 44.49 | Rubric Pass Rate (%) | 78.4 |
| AA-Briefcase - Analytical Quality Elo | 1514.24 | Elo | 74.2 |
| AA-Briefcase - PDF - Rubric Pass Rate | 45.69 | Rubric Pass Rate (%) | 74 |
| AA-Briefcase | 44.85 | Rubric Pass Rate (%) | 73 |
| AA-Briefcase - XLSX - Rubric Pass Rate | 42.27 | Rubric Pass Rate (%) | 71.6 |
| AutomationBench-AA | 48.78 | Guardrail-Adjusted Objective Completion (%) | 60.3 |
| AutomationBench-AA - Strict | 17.05 | Strict Task Success (%) | 60.1 |
| AutomationBench-AA - Finance - Raw Objective Completion | 75.66 | Raw Objective Completion (%) | 59.9 |
| AutomationBench-AA - Support - Raw Objective Completion | 80.11 | Raw Objective Completion (%) | 58.3 |
Interactive version: theaggregate.ai/model?slug=deepseek-v4-1-flash-1017-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-05.