Skip to content

DeepSeek · Chat

DeepSeek V3.1

DeepSeek 🇨🇳ChatDeepSeek familyOpen weights. One of its sources says the weights are publicly available.
From Aiml API's listing:
DeepSeek-V3.1 is the latest advanced LLM with improved architecture and performance across various natural language tasks.
llminfo IQ
119.6
Market median price
$0.6629 blended / 1M
Max context
164K
Listings
57

Evidence and sources

🤗 deepseek-ai/DeepSeek-V3.1llminfo IQ 119.6, ranked among chat models. From 122 benchmarks. Plenty of data behind it.How llminfo IQ is computedCoding IQ 94.3, 95% interval 86–103. From 9 benchmark families across 5 sources. A fair amount of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.How Coding IQ is computedPoor value (44.0). Costs more per task than its benchmark results predict. From 17 benchmarks with measured task costs. Plenty of data behind it.How Value for money worksReleased 21 Aug 2025 Training-data cutoff, as its sources state it Maker HQ: CN (according to DeepSeek's own website)
All rankings: #104 on OpenRouter today #124 on OpenRouter this week #125 on OpenRouter this monthSources: GenAI Prices data, Models.dev data, AI/ML API data, OpenRouter data docs, llminfo researchPrimary data from GenAI Prices from Models.dev: name, weights availability from AI/ML API: primary mode, description from OpenRouter: knowledge cutoff from llminfo research: release date

Advantages and trade-offs

Compared with 10 chat models, picked mainly by popularity and IQ

Advantages

Open weights available

The weights are public. 4 of the 9 peers that say either way have open weights.

Trade-offs

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 04:34 UTC.Prices come from the model's maker where they sell it, otherwise their official partner, otherwise the market median. This comparison uses 7 maker's prices; 3 market median prices.

Input & output

as stated by the model's providers
Text
Text

Capabilities

Supported operations and provider-stated features

ReasoningFunction callingStructured outputPrompt cachingBatch API

Listings

 OpenRouterdeepseek/deepseek-chat-v3.1:free · freeSource: Helicone Costs dataNot published by the provider$0.00$0.00
 Flowdotdeepseek/deepseek-chat-v3.1Source: FlowDot data33K$0.15$0.75
 Meganovadeepseek-ai/DeepSeek-V3.1 · fp16Sources: MegaNova data, Models.dev data164K$0.19$0.79
 Tokengodeepseek/deepseek-v3.1Source: Models.dev data131K$0.19cache read $0.06$0.71
 NanoGPTdeepseek-ai/DeepSeek-V3.1Sources: Models.dev data, NanoGPT128K$0.20cache read $0.10$0.70
 Submodeldeepseek-ai/DeepSeek-V3.1Source: Models.dev data75K$0.20$0.80
 Fastrouterdeepseek/deepseek-v3.1Source: FastRouter data164K$0.25cache read $0.13$0.95
 Hugging FaceInference by DeepInfradeepseek-ai/DeepSeek-V3.1Source: Hugging Face Inference Providers data164K$0.25$0.95
 Kilodeepseek/deepseek-chat-v3.1Sources: Kilo AI Gateway data, Models.dev data164K$0.25cache read $0.13$0.95
 Nous Portaldeepseek/deepseek-chat-v3.1Source: Nous Portal Inference Gateway data164K$0.25cache read $0.13$0.95

5 listings without a price are hidden.

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Science

GPQA Diamond8 results
Source: OpenRouter benchmark exports
deepseek-chat-v3.1 · 90-day results
Across 7 providers lowest 0.168% median 77.4% highest 83.2%OpenRouter's automatic routing 62.5%
Show results from 7 providers and OpenRouter's automatic routingHide results from 7 providers and OpenRouter's automatic routing
  • Novita
  • AtlasCloud
  • CoreWeave
  • DeepInfra
  • Mara
  • SambaNova
  • OpenRouter's automatic routing
  • SiliconFlow

Agentic

Source: OpenRouter benchmark exports
deepseek-chat-v3.1 · 90-day results
Across 4 providers lowest 30% median 44.8% highest 71.3%OpenRouter's automatic routing 43%
Show results from 4 providers and OpenRouter's automatic routingHide results from 4 providers and OpenRouter's automatic routing
  • Novita
  • SiliconFlow
  • OpenRouter's automatic routing
  • DeepInfra
  • CoreWeave

Reasoning

Data credits: OpenRouter public benchmark leaderboard and search-lane exports, CC BY 4.0. Fractional source scores converted to percentages; immutable dated snapshots. (licence)

Also known as: azure_ai/deepseek-v3.1, baseten/deepseek-ai/deepseek-v3.1, bedrock_mantle/ap-northeast-1/deepseek.v3.1, bedrock_mantle/ap-south-1/deepseek.v3.1, bedrock_mantle/ap-southeast-2/deepseek.v3.1, bedrock_mantle/ap-southeast-3/deepseek.v3.1, bedrock_mantle/deepseek.v3.1, bedrock_mantle/eu-west-2/deepseek.v3.1, bedrock_mantle/sa-east-1/deepseek.v3.1, cc-deepseek-v3.1, deepinfra/deepseek-ai/deepseek-v3.1, deepseek-ai/deepseek-v3-1, deepseek-ai/deepseek-v3.1, deepseek-ai/deepseek-v3.1-maas, deepseek-ai/deepseek-v3.1:thinking, deepseek-v3-1, deepseek-v3-1-250821, deepseek-v3.1, deepseek-v3.1-think, deepseek-v3.1-thinking, deepseek/deepseek-chat-v3.1, deepseek/deepseek-chat-v3.1:free, deepseek/deepseek-v3.1, fireworks_ai/accounts/fireworks/models/deepseek-v3p1, gmi/deepseek-ai/deepseek-v3.1, novita/deepseek/deepseek-v3.1, pro/deepseek-ai/deepseek-v3.1, replicate/deepseek-ai/deepseek-v3.1, sambanova/deepseek-v3.1, sophnet/deepseek-v3.1, sophnet/deepseek-v3.1-fast, streamlake/deepseek-v3.1, together_ai/deepseek-ai/deepseek-v3.1, vertex_ai/deepseek-ai/deepseek-v3.1-maas, wandb/deepseek-ai/deepseek-v3.1

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works