DeepSeek V4.1 Flash 1017
| Released | 2026-10-17 |
|---|---|
| Price per 1M | $0.15 in · $0.60 out |
| Overall score | 61.4 · rank 19 |
| Context window | 1,000,000 tokens |
| Provider | DeepSeek |
| Weights | Open weights (MIT) |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 61.4 | 19 | 34 |
| Overall: Coding | 62.5 | 12 | 26 |
| Overall: Reasoning | 60.8 | 28 | 17 |
| Overall: Agentic work | 62.2 | 18 | 24 |
| Overall: Frontend | 62.2 | 11 | 7 |
| Coding | 62.9 | 10 | 5 |
| Agents | 61.6 | 18 | 9 |
| Reasoning | 63.4 | 22 | 10 |
| Math | 60.6 | 26 | 5 |
| Knowledge | 55.4 | 56 | 2 |
| Frontend | 61.9 | 11 | 2 |
| General | 61.1 | 26 | 1 |
| Value vs price | +17.7 | 2 |
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 63.8 ±3.1 | 7 |
| High | 61.8 ±3.3 | 6 |
| Low | 62.0 ±4.0 | 3 |
| Max | 61.3 ±1.4 | 24 |
| Medium | 62.3 ±5.0 | 1 |
| Non-reasoning | 44.0 ±4.9 | 4 |
Evidence
Coding 62.9 ±1.8
| Benchmark | Result | Source | Date |
|---|---|---|---|
| CursorBench | 47.7% | Epoch AI | 2026-11-11 |
| Code Migration v1.1 | 67.5% | Vals AI | 2026-11-09 |
| LiveBench 2.0 Coding | 80.4% | LiveBench | 2026-11-06 |
| WeirdML v3 | 9.2% | Epoch AI | 2026-11-11 |
| SciCode | 55.9% | Epoch AI | 2026-11-11 |
Agents 61.6 ±1.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Terminal-Bench 4.0 (Vals run) | 31.8% | Vals AI | 2026-11-09 |
| APEX-Agents | 69.8% | Epoch AI | 2026-11-11 |
| LiveBench 2.0 Agentic | 38.8% | LiveBench | 2026-11-06 |
| LiveBench 2.0 Browser Use | 67.2% | LiveBench | 2026-11-06 |
| EMB v1.1 (Excel financial modeling) | 69.2% | Vals AI | 2026-11-09 |
| LiveBench 2.0 Research | 78.6% | LiveBench | 2026-11-06 |
| Tax Agent Bench | 44.4% | Vals AI | 2026-11-09 |
| Finance Agent v2.1 | 62.9% | Vals AI | 2026-11-09 |
| HLAB-Vals 1.1 (legal agent) | 14.2% | Vals AI | 2026-11-09 |
Reasoning 63.4 ±1.5
| Benchmark | Result | Source | Date |
|---|---|---|---|
| ARC-AGI-2 | 78.1% | Epoch AI | 2026-11-11 |
| Extended NYT Connections | 92.8% | Lech Mazur | 2026-11-10 |
| ARC-AGI-2 (ARC Prize verified) | 78.1% | ARC Prize | 2026-11-11 |
| GPQA Diamond | 93.2% | Epoch AI | 2026-10-19 |
| CritPt | 24.0% | Epoch AI | 2026-11-11 |
| ARC-AGI-1 (ARC Prize verified) | 94.5% | ARC Prize | 2026-11-11 |
| ARC-AGI-1 | 94.5% | Epoch AI | 2026-11-11 |
| ARC-AGI-3 | 5.8% | ARC Prize | 2026-11-11 |
| Chess Puzzles | 46.0% | Epoch AI | 2026-10-19 |
| LiveBench 2.0 Reasoning | 89.1% | LiveBench | 2026-11-06 |
| Mystery Game Puzzles | 49.0% | Epoch AI | 2026-10-19 |
| LiveBench 2.0 Data Analysis | 79.5% | LiveBench | 2026-11-06 |
Math 60.6 ±1.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| FrontierMath Tiers 1-3 | 71.2% | Epoch AI | 2026-10-19 |
| FrontierMath Tier 4 | 46.3% | Epoch AI | 2026-10-19 |
| MathArena ArXivMath (latest month) | 38.3% | MathArena | 2026-11-11 |
| OTIS Mock AIME 2024-2025 | 99.4% | Epoch AI | 2026-10-19 |
| MathArena BrokenArXiv (latest month) | 49.7% | MathArena | 2026-11-11 |
Knowledge 55.4 ±2.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| SimpleQA Verified | 42.5% | Epoch AI | 2026-10-19 |
| Legal Research Bench | 45.2% | Vals AI | 2026-11-09 |
Frontend 61.9 ±0.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Vibe Code Bench v1.1 | 87.7% | Vals AI | 2026-11-09 |
| LiveBench 2.0 Design | 79.8% | LiveBench | 2026-11-06 |
General 61.1 ±3.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LLM Debate Benchmark | 1,593 | Lech Mazur | 2026-10-26 |
Results not used in the score
| Benchmark | Result | Source | Why not used |
|---|---|---|---|
| ARC-AGI-1 | 93.5% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 | 91.5% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 | 5.7% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 93.5% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 91.5% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 5.7% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 | 68.6% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 | 73.6% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 | 2.1% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 68.6% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 73.6% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 2.1% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-3 | 0.2% | ARC Prize | A stronger configuration result was used |
| CritPt | 0.3% | Epoch AI | A stronger configuration result was used |
| CursorBench | 46.3% | Epoch AI | A stronger configuration result was used |
| CursorBench | 45.6% | Epoch AI | A stronger configuration result was used |
| CursorBench | 44.9% | Epoch AI | A stronger configuration result was used |
| SciCode | 38.5% | Epoch AI | A stronger configuration result was used |
Names matched to this model
deepseek-flash, deepseek-v4-1-flash-1017