AI Observatory / Model Radar DeepSeek / deepseek/deepseek-chat-v3.1

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...

32,768 Context
7,168 Max output
$0.15 / 1M Prompt price
2025-08-21 Created
01 / Snapshot

Pricing, context, modalities, and parameters.

Model Radar detail pages stay neutral and operator-readable: core metadata first, then workflow fit.

Provider DeepSeek Input modalities text
Output modalities text Prompt price $0.15 / 1M
Completion price $0.75 / 1M Request price N/A
Context length 32,768 Max completion tokens 7,168
Supported parameters frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, min_p, presence_penalty, reasoning, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p
Best for deepseek/deepseek-chat-v3.1

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...

Deep analysis High-volume usage Coding workflows tool-capable low-cost
03 / Colophon

Routes and exits.

Each model page stays simple: overview, compare, related models, then back to the public hub.