The Model Index
Models/DeepSeek

DeepSeek-V3.2

DeepSeekReleased 1 Dec 2025Open
CompareAnnouncement
145134–156Index score#70 of 229 models
3/ 8Core tests taken3 of 6 areas · plus 11 anchor tests
−16Versus the frontier at releasebest then: Gemini 3 Pro
$0.28/ $0.42Price per million tokensinput / output, list price
164KContext windowtokens it reads at once

Where it sitson the frontier

o3 reached this score 8 months earlier.

Test resultsby area · tick = best by any model

Reasoning & knowledgeevidence
GPQA Diamond
83%
Humanity's Last Exam
no result
Mathsevidence
FrontierMath T1-3
no result
Codingevidence
SWE-bench Verified
no result
Terminal-Bench
40%
Agentsevidence
APEX-Agents
no result
Novel problemsevidence
ARC-AGI-2
4%
Long tasksno evidence yet
METR time horizon
no result

Independent results from the Epoch AI benchmarking hub, best reasoning setting. The core tests are shown here; anchor tests also inform the score and can be shown above.

What others sayrank on each leaderboard

The Model Indexindependent tests, our method#70of 219 · 145
LMArena · Creative writingcreative writing#95of 386 · 1401LMArena · Mathsmaths questions in chat#97of 373 · 1428LMArena · Codingcoding questions in chat#100of 383 · 1476LMArena · Textoverall chat quality, judged by people#107of 388 · 1425Epoch Capabilities Indexoverall capability from many benchmarks#70of 239 · 146.3LMArena · WebDevbuilding web apps and front ends#89of 123 · 1362

Further left is better. Leaderboards measure different things: LMArena is people voting blind between two answers, the others are test scores. Click any row for the source.

Best attop three among current models

Not in the top three among current models on any category we track.

Where this model places in the top three of today's models, on category leaderboards and on our core tests.

Where to use itprices checked 1 Oct 2026

From the lab

DeepSeek Chatapp · subscription↗DeepSeek APIAPI↗

API providers 13 · per million tokens, in / out

GGMICloud$0.21 / $0.31 · fp8cheapestDDeepInfra$0.26 / $0.38 · fp4↗AAtlasCloud$0.26 / $0.38 · fp8↗VVenice$0.27 / $0.39↗SSiliconFlow$0.26 / $0.42 · fp8↗BBaidu$0.28 / $0.42 · fp8↗DDigitalOcean$0.30 / $0.96↗AAlibaba$0.37 / $1.11 · fp8↗

Run it yourself

🤗Official weightsdeepseek-ai/DeepSeek-V3.2↗

Provider prices come from OpenRouter's live listing; going direct to a provider can differ. Subscriptions are the lab's own apps.

Facts and sources

Released1 Dec 2025
Release entryDeepSeek-V3.2 (+ Speciale)
Scored asDeepSeek-V3.2
Artificial Analysis index–
Output speed–

DeepSeek releasesaround this one

DeepSeek-V3.2-ExpOpen
index 143
DeepSeekMath-V2Open
DeepSeek-V3.2Open
index 145
DeepSeek-V4-Flash (Apr 2026)Open
index 145
DeepSeek-V4-Pro (Apr 2026)Open
index 151