Deepseek v3

DeepSeekdeepseek/deepseek-v3
Modality

Chat

Family

DeepSeek

Context window

—

Max output

164K tokens

Served On

Amazon Bedrock

Deepseek v3 is a chat model made by DeepSeek, served on nRouter.

Model pricing

Exact provider list price · 0% platform fee · x-nr-request-cost header

Rates are read live

Pricing for deepseek/deepseek-v3 is fetched from the live catalogue on load, so it is never served from a cache that could outlive a repricing.

Start building with deepseek/deepseek-v3

One API key, three steps. Choose your tool to begin.

  1. Set your API key

    export NROUTER_API_KEY="<YOUR_API_KEY>"
  2. Configure Claude Code

    Choose a tool, then copy its setup from the panel.

npm install -g @anthropic-ai/claude-code
Claude Code · deepseek/deepseek-v3
# Point Claude Code to nRouter gateway
export ANTHROPIC_BASE_URL="https://api.nrouter.ai"
export ANTHROPIC_API_KEY="$NROUTER_API_KEY"

# Launch with deepseek/deepseek-v3
claude --model deepseek/deepseek-v3

Specifications & capabilities

VisionFunction callingSystem prompt
Compare context window against peer models (—)

Context window, ranked against other chat models

Proportional scale: 0 to 272K tokens. Not ranked, no published context window: alibaba/qwen-mt-flash, alibaba/qwen3.7-flash, google/gemini-4-neon.

Benchmarks & performance rankings

Published LLM Arena category placements and Elo ratings for Deepseek v3.

LLM Arena category ratings
Overall#234

1333Elo

Coding#247

1325Elo

Math#237

1311Elo

Creative Writing#202

1329Elo

Hard Prompts#248

1312Elo

Instruction Following#229

1315Elo

Multi-Turn#214

1349Elo

Longer Query#215

1343Elo

Non-English#228

1316Elo

Source: huggingface.co · CC-BY-4.0

Availability

Collecting probe samples

Not enough health probe samples collected yet to publish an uptime or latency figure.

More from DeepSeek

FAQs

What is Deepseek v3?

Deepseek v3 is a chat model made by DeepSeek, served on nRouter.

How much does Deepseek v3 cost?

Deepseek v3 is billed at the provider's list price on nRouter with no per-token markup — tokens settle at the exact upstream rate. Input and output rates are read live, so see the pricing section above for the current figures.

Does Deepseek v3 support tool calling and structured outputs?

No. Deepseek v3 does not support tool calling, so it cannot invoke external tools or functions.

What inputs and outputs does Deepseek v3 support?

Deepseek v3 takes text as input and returns text as output.

How do I use Deepseek v3 via nRouter?

Use any OpenAI-compatible client: point it at https://api.nrouter.ai/v1 with your nRouter API key and set the model to "deepseek/deepseek-v3". The quick-start panel above has a copy-paste curl command for this model.

Where is the Deepseek v3 documentation?

Start at nrouter.ai/docs. The quick-start guide (nrouter.ai/docs/getting-started/quick-start) covers API keys and your first request, and the API reference (nrouter.ai/docs/api-reference) lists every parameter.

What other text models does DeepSeek have?

Other DeepSeek text models on nRouter include DeepSeek V3.2 (nrouter.ai/models/deepseek-deepseek-v3.2), DeepSeek V4 Pro 0813 (nrouter.ai/models/deepseek-deepseek-v4-pro-0813), DeepSeek V4.1 Flash (nrouter.ai/models/deepseek-deepseek-v4.1-flash), DeepSeek V4 Pro (nrouter.ai/models/deepseek-deepseek-v4-pro).