Models / DeepSeek V3

DeepSeek V3

DeepSeek License

DeepSeek's flagship MoE model. 671B total with 37B active. Competitive with GPT-4o on coding and reasoning.

deepseek/deepseek-v3
Context Window
131K
Max Output
16K
Providers
2
Released
2025-03

Capabilities

chatcodereasoningtoolsstreamingjson_mode

Pricing by Provider

ProviderInput $/1MOutput $/1MLatency p50Latency p95Status
deepseek$0.27$1.10400ms1200ms
together$0.30$1.20450ms1300ms

Quick Start

Python
import magicrouter

mr = magicrouter.Client(
    provider_keys={"deepseek": "your-api-key"}
)

response = mr.chat(
    "deepseek/deepseek-v3",
    "Your prompt here"
)
print(response.choices[0].message.content)
TypeScript
import { MagicRouter } from "magicrouter";

const mr = new MagicRouter({
  providerKeys: { deepseek: "your-api-key" }
});

const response = await mr.chat({
  model: "deepseek/deepseek-v3",
  messages: [{ role: "user", content: "Your prompt here" }]
});
console.log(response.choices[0].message.content);
cURL
curl https://api.deepseek.com/v1/chat/completions \
  -H "Authorization: Bearer your-api-key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-chat",
    "messages": [{"role": "user", "content": "Your prompt here"}]
  }'

Use this model

Sign up for free and test DeepSeek V3 in the playground

Get Started
DeepSeek V3 — MagicRouter