AiCostCompare

DeepSeek V3.1

by DeepSeek · deepseek/deepseek-chat-v3.1 · #108 cheapest of 325 paid models

Prices updated Jul 22, 2026, 2:49 AM UTC · refreshed hourly

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...

Compare vs:Claude Sonnet 5Claude Opus 4.6Claude Haiku 4.5GPT-5.6 Sol

Pricing

Input

$0.25

per 1M tokens

Output

$0.95

per 1M tokens

Blended (3:1)

$0.425

3 input : 1 output

Cache read

$0.13

per 1M cached tokens

Cache write

per 1M tokens

Cache & batch economics

Effective prices for DeepSeek V3.1 with prompt caching.

Cache hit rateEffective blended $/1MExample request*vs no cache
0%$0.425$0.0265
50%$0.38$0.0205−23%
90%$0.344$0.0157−41%

*Example: 100K sticky context + 2K new input + 1K output. Batch discount is an approximate 50% off list rates (available for OpenAI / Anthropic / Google — not flagged for this creator). Confirm on the vendor’s pricing page.

What a request costs

WorkloadInput tokensOutput tokensCost / requestCost / 1K requests
Short chat message500300$0.00041$0.41
Document summary8,0001,000$0.00295$2.95
Codebase question (RAG)30,0002,000$0.0094$9.40
Long-context analysis150,0005,000$0.0422$42.25

Performance

Intelligence Index

21

AA composite quality

Coding Index

Math Index

49.7

Agentic Index

Output speed

0 tok/s

median

Time to first token

0s

median

Time to first answer

0s

after reasoning tokens

Artificial Analysis benchmarks

BenchmarkScore
MMLU-Pro83.3%
GPQA Diamond73.5%
Humanity's Last Exam6.3%
LiveCodeBench57.7%
SciCode36.7%
AIME 202549.7%
IFBench37.8%
AA-LCR (Long Context Reasoning)45.0%
Terminal-Bench Hard24.2%
τ²-Bench (Telecom)34.8%

Design Arena Elo

CategoryElo
Website1,150
UI components1,126
Data viz1,136
SVG1,014
3D1,136
Game dev1,143
Code categories1,145

Specs

Context window

164K

tokens

Max output

33K

tokens

Input modalities

text

Tokenizer

DeepSeek