LLM Reference

DeepSeek V3.1

Released
2025-08-21
Last refreshed
2026-06-29
Status
Researched 56d ago
Open sourceCommercial use: permittedMultimodalCodingAgentsVisionClassificationJSON / Tool use

DeepSeek V3.1 is worth evaluating for coding, agents, and vision when its provider route and context window match the workload.

Use it for

  • Teams evaluating coding, agents, and vision
  • Workloads that can use a 64k context window
  • Buyers comparing 4 tracked provider routes

Do not use it for

  • Workloads where another current model has stronger sourced task evidence
Specifications
Released
2025-08-21
Context
64k
Parameters
671B total, 37B active (MoE)
Architecture
Mixture of Experts
Specialization
general
Openness
Open source
License
MITOSI-approvedCommercial use: permitted
Weights
Unknown
Code
Unknown
Training
Fine-tuned
Created by

Advancing artificial general intelligence (AGI).

Hangzhou, Zhejiang, China
Founded 2023
Website
Pricing
Output / 1M
$1.00
Input / 1M
$0.270

Cheapest of 8 routes · Novita AI

About

Enhanced reasoning and grounded retrieval model from DeepSeek with multimodal text and image understanding.

DeepSeek V3.1 is an open-source model in the DeepSeek V3 family. The structured metadata tracks a 64k-token context window, multimodal input, structured outputs, and code execution. This page tracks provider routes through Microsoft Foundry, Fireworks AI, NVIDIA NIM, and 5 more, with the cheapest tracked route listed at $0.27 input and $1 output per 1M tokens. Headline tracked benchmarks include SWE-bench Verified 66.0 and MMLU PRO 83.3.

Top use-case fit: coding, agents, and build tasks

Coding

Q/$ C

1 relevant benchmark in the decision map.

Agents

Q/$ B

1 relevant benchmark in the decision map.

Vision

Included by capability and metadata signals in the decision map.

Provider price ladder

Compare all 8

Compare API pricing across 4 providers for input and output tokens, batch, and cached reads when available.

ProviderInput / 1MOutput / 1MCacheRoute
Novita AI$0.270$1.00-
Serverless
Fireworks AI$0.560$1.68-
Serverless
Vercel AI Gateway$0.560$1.68read $0.280
Serverless
Together AI$0.600$1.70-
Serverless

Available via routers & gateways(8)

Capabilities

VisionMultimodalStructured OutputsCode Execution

Benchmark peer barsfor Coding

Benchmark scores(2)

Scores are benchmark-specific and are direction-aware: the same numeric gap can mean very different outcomes across suites. Use the leaderboard context and this model's provider route to decide whether the winning margin is meaningful for your workload.
BenchmarkScoreVersionSource
SWE-bench Verified66.02025https://www.swebench.com/
MMLU PRO83.3https://huggingface.co/spaces/TIGER-Lab/MMLU-Pro

Migration checks

No linked migration route is available for this model yet.

Compare DeepSeek V3.1 with other models

Show all 50 popular comparisonssorted by 7-day search impressions
DeepSeek V3.1 vs Gemini 2.5 Pro138DeepSeek V3.1 vs Kimi K2.5132DeepSeek V3.1 vs Llama 3.1 405B Instruct100DeepSeek V3.1 vs Tencent Hunyuan Turbo S92DeepSeek V3.1 vs GPT-5.5 Pro79DeepSeek V3.1 vs Claude Opus 4.774DeepSeek V3.1 vs Xiaomi MiMo-V2.5-TTS-Series68DeepSeek V3.1 vs Qwen3.6-27B66DeepSeek V3.1 vs o361DeepSeek V3.1 vs Claude Sonnet 4.653DeepSeek V3.1 vs Qwen3.6-35B-A3B51DeepSeek V3.1 vs Gemini 2.5 Flash51DeepSeek V3.1 vs GPT-5.2 Codex50DeepSeek V3.1 vs Claude Sonnet 4.548DeepSeek V3.1 vs DeepSeek V332DeepSeek V3.1 vs GPT-5.232DeepSeek V3.1 vs Qwen3.5-9B29DeepSeek V3.1 vs Mistral Large 226DeepSeek V3.1 vs Qwen3.5-397B-A17B25DeepSeek V3.1 vs Llama 3.2 1B21DeepSeek V3.1 vs Claude 3.7 Sonnet18DeepSeek V3.1 vs GPT-4 Vision Preview18DeepSeek V3.1 vs Claude Opus 4.618DeepSeek V3.1 vs Mixtral 8x22B Instruct v0.316DeepSeek V3.1 vs GPT-5.415DeepSeek V3.1 vs Llama 3 8B Instruct15DeepSeek V3.1 vs Mixtral 8x7B14DeepSeek V3.1 vs o3 Mini14DeepSeek V3.1 vs Llama 3.1 70B Instruct14DeepSeek V3.1 vs Llama 3 70B Instruct11DeepSeek V3.1 vs Phi-4 Reasoning Vision 15B9DeepSeek V3.1 vs Mistral Magistral Small 25098DeepSeek V3.1 vs GPT-5.5 Instant7DeepSeek V3.1 vs Code Cushman 0027DeepSeek V3.1 vs Together AI Qwen2-7B-Instruct6DeepSeek V3.1 vs GPT-4o-mini Search Preview6DeepSeek V3.1 vs GPT-4 Turbo5DeepSeek V3.1 vs Qwen2-7B-Instruct5DeepSeek V3.1 vs o3-pro5DeepSeek V3.1 vs GLM-5V-Turbo4DeepSeek V3.1 vs Qwen3-9B4DeepSeek V3.1 vs Phi-4 Mini Flash Reasoning4DeepSeek V3.1 vs Claude Opus 4.54DeepSeek V3.1 vs Llama 3.2 1B Instruct3DeepSeek V3.1 vs Code Davinci 0013DeepSeek V3.1 vs ShieldGemma 9B2DeepSeek V3.1 vs GPT-5.5-Cyber2DeepSeek V3.1 vs Together AI - Llama 3 8B Lite2DeepSeek V3.1 vs Mistral Nemotron2DeepSeek V3.1 vs o3 Deep Research2

Frequently asked questions

What is the context window of DeepSeek V3.1?

DeepSeek V3.1 has a context window of 64k tokens.

How much does DeepSeek V3.1 cost?

DeepSeek V3.1 pricing ranges from $0.27/1M to $0.672/1M input tokens depending on the provider.

When was DeepSeek V3.1 released?

DeepSeek V3.1 was released on 2025-08-21.

Which providers offer DeepSeek V3.1?

DeepSeek V3.1 is available from 8 providers: Microsoft Foundry, Fireworks AI, NVIDIA NIM, Together AI, AWS Bedrock, Replicate API, Vercel AI Gateway, Novita AI.

What benchmarks has DeepSeek V3.1 been tested on?

DeepSeek V3.1 has been evaluated on 2 benchmarks, including SWE-bench Verified, MMLU PRO.