deepseek
DeepSeek: DeepSeek V3.1
DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context...
Context
164k tokens
Input / 1M tokens
$0.21
Output / 1M tokens
$0.79
Benchmark
—
Pricing
Input tokens$0.21 per 1M tokens
Output tokens$0.79 per 1M tokens
Cache read$0.13 per 1M tokens
Technical details
Model IDdeepseek/deepseek-chat-v3.1
Context window164k tokens
Input modalitiestext
Output modalitiestext
TokenizerDeepSeek
Max output tokens32,768
Use with MegaBrain
import OpenAI from 'openai'
const client = new OpenAI({
baseURL: 'https://getmegabrain.com/api/gateway/v1',
apiKey: process.env.MEGABRAIN_API_KEY,
})
const response = await client.chat.completions.create({
model: 'deepseek/deepseek-chat-v3.1',
messages: [{ role: 'user', content: 'Hello!' }],
})Ready to use DeepSeek: DeepSeek V3.1?
Get an API key and start making requests in minutes.
Get an API key