DeepseekTool use

DeepSeek V3.1 pricing

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 training process, reaching up to 128K tokens, and uses FP8 microscaling for efficient inference. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default. Live index: 1 priced offer. Best input $0.200 per million tokens from Openrouter. Best output $0.800 per million tokens from Openrouter.

164K context·1 provider·verified May 2, 2026
Best input$0.200per 1M tokens · Openrouter
Best output$0.800per 1M tokens · Openrouter

Pricing across providers

Every row is a seller of DeepSeek V3.1 with token pricing we track. The cheapest input in this snapshot is from Openrouter. The bar chart shows the same input and output dollars per million for a quick scan.

O
Openrouter
Input / 1M
$0.200
Output / 1M
$0.800

Input vs output · 1M tokens

Cost calculator

The calculator uses the same dollars per million tokens as the table. Adjust sliders to see how DeepSeek V3.1 cost scales with traffic.

In: $0.200/M·Out: $0.800/M

0.020000¢ / req

0.040000¢ / req

Daily
$6.00
Monthly
$180
Annual
$2.2K

Model specifications

These fields describe DeepSeek V3.1 as we store it (source: Deepseek). They sit next to price so buyers can check limits and tools in one place.

Context window
163,840 tokens
Max output
163,840 tokens
Vision (images)
No
Tool / function calling
Yes
Streaming
No
Released
Aug 2025
Primary provider
Deepseek
Model family
N/A

Compare DeepSeek V3.1

These links open full side by side pages for DeepSeek V3.1. We picked pairs that people often shop together. 6 ready to open.

Frequently asked questions

Read these after the table if you want plain language around DeepSeek V3.1 rates. The short model note from our index: 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 trainin...

Yes. DeepSeek V3.1 is available on Openrouter.

Also from Deepseek

Other models by Deepseek with live pricing in our catalog.