# DeepSeek V3.1 Benchmark Scores & Performance

> DeepSeek V3.1 by DeepSeek scores 40.78/100 overall, ranking #137 out of 954 AI models.

Canonical page: https://benchlm.ai/models/deepseek-v3-1

Last updated: October 17, 2026

## Model Details

| Property | Value |
|----------|-------|
| Creator | DeepSeek |
| Source Type | Open Weight |
| Reasoning Type | Non-Reasoning |
| Context Window | 128K |
| Overall Score | 40.78/100 |
| Overall Rank | #137 of 954 |

## Family & Coverage

- Family: DeepSeek V3.1
- Variant: base
- Benchmarks covered: 17 of 639
- Sibling models: [DeepSeek V3.1 (Reasoning)](/models/deepseek-v3-1-reasoning)
- Coverage note: BenchLM currently has partial benchmark coverage for this model, so the overall score is conservative.

## Agentic Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [τ²-bench results](/benchmarks/tau2-bench) | 34.8% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/deepseek-v3-1) |

## Coding Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [SWE-bench Verified](/benchmarks/swe-bench-verified) | 66% | [DeepSeek-V3.1 model card](https://huggingface.co/deepseek-ai/DeepSeek-V3.1) |
| [LiveCodeBench](/benchmarks/livecodebench) | 56.4% | [DeepSeek-V3.1 model card](https://huggingface.co/deepseek-ai/DeepSeek-V3.1) |

## Multimodal & Grounded Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [Design Arena Website](/benchmarks/designarenawebsite) | 1124 | [OpenRouter model benchmarks](https://openrouter.ai/deepseek/deepseek-chat-v3.1/benchmarks) |

## Reasoning Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [AA-LCR](/benchmarks/lcr) | 47.0% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/deepseek-v3-1) |
| [CritPt](/benchmarks/critpt) | 0.0% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/deepseek-v3-1) |

## Knowledge Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [GPQA](/benchmarks/gpqa) | 74.9% | [DeepSeek-V3.1 model card](https://huggingface.co/deepseek-ai/DeepSeek-V3.1) |
| [MMLU-Pro](/benchmarks/mmlu-pro) | 83.7% | [DeepSeek-V3.1 model card](https://huggingface.co/deepseek-ai/DeepSeek-V3.1) |
| [Artificial Analysis Intelligence Index](/benchmarks/artificialanalysis) | 13.7% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/deepseek-v3-1) |
| [AA-GPQA Diamond](/benchmarks/gpqa-diamond) | 73.5% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/deepseek-v3-1) |
| [AA-HLE](/benchmarks/aahle) | 6.7% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/deepseek-v3-1) |
| [AA-Omniscience Index](/benchmarks/aaomniscienceindex) | -42.7% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/deepseek-v3-1) |
| [AA-Omniscience Accuracy](/benchmarks/omniscienceaccuracy) | 23.1% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/deepseek-v3-1) |
| [AA-Omniscience Hallucination Rate](/benchmarks/omnisciencehallucinationrate) | 85.7% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/deepseek-v3-1) |

## Instruction Following Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [AA-IFBench](/benchmarks/ifbench) | 37.8% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/deepseek-v3-1) |

## Mathematics Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [AIME 2024](/benchmarks/aime2024) | 66.3% | [DeepSeek-V3.1 model card](https://huggingface.co/deepseek-ai/DeepSeek-V3.1) |
| [AIME 2025](/benchmarks/aime2025) | 49.8% | [DeepSeek-V3.1 model card](https://huggingface.co/deepseek-ai/DeepSeek-V3.1) |

## Other DeepSeek Models

- [DeepSeek V4.1 Flash 1017](/models/deepseek-v4-1-flash) - Score: 65.61
- [DeepSeek V4 Pro 0813](/models/deepseek-v4-pro-0813) - Score: 63.19
- [DeepSeek V3.2](/models/deepseek-v3-2) - Score: 50.2
- [DeepSeek-R1](/models/deepseek-r1) - Score: 41.08
- [DeepSeek V3.1 (Reasoning)](/models/deepseek-v3-1-reasoning) - Score: 39.18
- [DeepSeek V3 0324](/models/deepseek-v3-0324) - Score: 36.5
- [DeepSeek V3](/models/deepseek-v3) - Score: 31.9
- [DeepSeek R1 Distill Qwen 32B](/models/deepseek-r1-distill-qwen-32b) - Score: 30.81
- [DeepSeek V4 Flash 0731](/models/deepseek-v4-flash-0731) - Score: not computed
- [DeepSeek-R1-0528](/models/deepseek-r1-0528) - Score: not computed
- [DeepSeek V3.2 (Thinking)](/models/deepseek-v3-2-thinking) - Score: not computed
- [DeepSeek LLM 2.0](/models/deepseek-llm-2-0) - Score: not computed
- [DeepSeek Coder 2.0](/models/deepseek-coder-2-0) - Score: not computed
- [DeepSeekMath V2](/models/deepseekmath-v2) - Score: not computed
- [DeepSeek-V2.5](/models/deepseek-v2-5) - Score: not computed