DeepSeek · deepseek/deepseek-chat-v3.1

deepseek-chat-v3.1

Budget Everyday 164k context Hermes Limited

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...

Input / M

$0.250

Output / M

$0.950

Value score

77.17

30d price Δ

Reasoning (AA)

13.7

Coding (AA)

—

Tool / agentic

—

Knowledge cutoff

2025-03-31

Price history

Not enough history yet — newer than the backfill window.

Capability snapshot

Last-known Artificial Analysis indexes via OpenRouter.

Index	Score
context	40.3
general_intelligence	13.7
reasoning	13.7