DeepSeek V3.1

Aug 2025
DeepSeek V3.1
Compare in
on
—per task?
$0.25 / $0.95per 1M in / out?
98 tok/sfirst token in 1.4 s?
128Kcontext

Rankingin non-thinking mode?

Strong at

C
Tool calling#134 of 242Reasoning#231 of 336Science#162 of 312Long context#183 of 314Knowledge#183 of 320Instructions#168 of 245

Alternatives?

ModelScoreCostPer taskWhat changes
CheaperGPT OSS 120BhighOpen+88%-86%$0.11BetterGPT OSS 20BhighOpen+80%-92%$0.011OpenDeepSeek V3.2 Specialethinking+95%——CheaperNemotron 3 Nano 30B A3BthinkingOpen+83%-75%$0.017CheaperSeed OSS 36B InstructthinkingOpen+70%-29%$0.21 / $0.57

Run it yourself

estimate
Fits on✗ 8 GB (phone, laptop)✗ 16 GB✗ 24 GB GPU✗ 48 GB GPU✗ 80 GB GPU
LicenceMITfree for commercial use

History

5 versions
20406080100Jan 25Apr 25Jul 25Oct 25Release dateAIME 2025 · AA archivedV3 · DecV3 · MarV3.1V3.2
20406080100Jan 25Jul 25Release dateAIME 2025 · AA archivedV3 · DecV3 · MarV3.1V3.2

Benchmarks

all categories
Chat & Knowledge
Humanity's Last Exam6.7#233
AA Omniscience Index-42.8#184
CritPt (physics)0#181
AA Intelligence Index 4.3.2≈13.7#159
AA Omniscience · Accuracy23.1#130
AA Omniscience · Hallucination Rate85.7#212
AA-LCR (long context) 1.147#185
GPQA Diamond73.5#164
AIME 2025 · AA archived49.7#89
Coding
Terminal-Bench Hard (version unstated)24.2#76
LiveCodeBench · AA archived57.7#77
Agents & Tool Use
τ²-bench · AA34.8#136
Writing & Language
IFBench (instructions)37.8#170

FAQ

How much does DeepSeek V3.1 cost?›

$0.25 per 1M input tokens and $0.95 per 1M output tokens (cheapest host: DeepInfra).

Is DeepSeek V3.1 free to use?›

There is no free API; the cheapest paid price is $0.25 per 1M input tokens. The weights are open, so you can run it yourself.

How good is DeepSeek V3.1, and what is it best at?›

#231 of 336 in Chat & Knowledge, #76 of 239 in Coding, #134 of 242 in Agents & Tool Use, #65 of 197 in Writing & Language.

Is DeepSeek V3.1 better than DeepSeek V3?›

On AIME 2025 · AA archived, DeepSeek V3.1 scores 89.7 vs 26 — higher.

How can I use DeepSeek V3.1?›

Through the API of AWS Bedrock, CoreWeave, DeepInfra, Featherless, SambaNova and 1 more.

What is the context window of DeepSeek V3.1?›

128K tokens, with up to 33K output tokens.

When was DeepSeek V3.1 released? Is there a newer version?›

Released 2025-08-21. It has been superseded by DeepSeek V3.1 Terminus.

Is DeepSeek V3.1 open source?›

Yes, the weights are open (MIT licence).

Can I run DeepSeek V3.1 locally?›

The weights are open and can be self-hosted; hardware needs are not confirmed yet.

Can DeepSeek V3.1 read images?›

No, it accepts text only.

What is DeepSeek V3.1?›

DeepSeek V3.1 is an open-weights language model from DeepSeek, released Aug 2025, the successor to DeepSeek V3.

What are the best alternatives to DeepSeek V3.1?›

GPT OSS 120B (cheaper), GPT OSS 20B (better), DeepSeek V3.2 Speciale (open weights).