Skip to content
AI Atlas
ModelAnnouncedOpen weightsIdentity probable

DeepSeek V3.2 Exp

DeepSeekfamily · DeepSeek

DeepSeek-V3.2-Exp is an experimental large language model released by DeepSeek as an intermediate step between V3.1 and future architectures. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism...

Open in Graph
data quality67

Updated 5 d ago · first seen 11 Sept 2026

model_01M294WWB78928015NA9WBEVMF

Overview

Identity

Canonical model
Yesidentity confidence: mediumOne row per real model release. Artifacts (checkpoints, quantisations, conversions) and folded evaluation variants point here.
Official checkpoints
official_checkpoints = hf_repo identifiers carried by the model itself; artifacts are separate entities pointing here through canonical_id.
Artifacts
None recordedSeparate entities (checkpoint · quantization · conversion · packaging) pointing to this model through canonical_id.
Provider deployments
2
API aliases
deepseek-v3-2-reasoning-0925deepseek/deepseek-v3.2-expIdentifiers under which providers and evaluators refer to this model.
Folded evaluation variants
0Effort / thinking variants (…-high, …-non-reasoning) are result configurations of this model, not separate models. Their old URLs redirect here.

Openness

Open weights— weights downloadable; 7 dimensions unknown.

Weights downloadable under a permissive or Creative Commons licence allowing commercial use; code or data may be missing.

  • Weights

    Yes

  • Inference code

    —

  • Training code

    —

  • Training data

    —

  • Dataset

    —

  • Commercial use

    —

  • Redistribution

    —

  • Derivatives

    —

dimensions marked null are unknown, not false

Key facts

Release date

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Status

Source:DeepSeek — site & API docsT2observed 5 d agomediumLLM-extracted

Version

Source:DeepSeek — site & API docsT2observed 5 d agomediumLLM-extracted

Knowledge cutoff

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Paper
github.com/deepseek-ai/DeepSeek-V3.2-Exp/blob/main/DeepSeek_V3_2.pdf

Source:DeepSeek — site & API docsT2observed 5 d agomediumLLM-extracted

Repository
huggingface.co/deepseek-ai/DeepSeek-V3.2-Exp

Source:DeepSeek — site & API docsT2observed 5 d agomediumLLM-extracted

Openrouter id

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Architecture

Architecture

Source:DeepSeek — site & API docsT2observed 5 d agomediumLLM-extracted

Tokenizer

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Hugging Face repo

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Capabilities

Modalities

Modalities
text
Input
text
Output
text

Capabilities

  • Tool calling

    Yes

    OpenRouter public model & pricing listing · T2

  • Structured output

    Yes

    OpenRouter public model & pricing listing · T2

  • Reasoning

    Yes

    OpenRouter public model & pricing listing · T2

  • Vision

    Unavailable

  • Audio

    Unavailable

  • Fine-tuning available

    Unavailable

Context window

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Max output

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Knowledge cutoff

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Tokenizer

Source:OpenRouter public model & pricing listingT2observed 5 d agomedium

Benchmarks18

Compare with another model →

Comparable same task and conditions · Partially comparable same task, conditions differ (effort, temperature, judge) · Not comparable different variant or metric

Benchmark results grouped by comparability group
Benchmark · groupBest scoreTrustConfigurationResultsvs leaderEvaluatedSource
Terminal-Benchagentic · accuracy · variant=hard · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysisvarianthardreasoningonconditions differ across rows → partially comparable3−34.8 ptvs gpt-5.6-solobs. 16 Sept 2026artificialanalysis.aiT2
τ²-benchagentic · pass^1 · variant=Telecom · evaluator=Artificial AnalysisIndependentevaluatorArtificial AnalysisvariantTelecomreasoningoffconditions differ across rows → partially comparable2non-primary groupobs. 16 Sept 2026artificialanalysis.aiT2
τ²-benchagentic · pass^1 · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysis1−65.2 ptvs Z.ai GLM 5.2obs. 11 Sept 2026artificialanalysis.aiT2
Artificial Analysis Intelligence Indexcomposite · indexIndependentreasoningonversion4.3conditions differ across rows → partially comparable3−36.8vs Claude Fable 5.1obs. 16 Sept 2026artificialanalysis.aiT2
IFBenchinstruction-following · accuracy · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysisreasoningonconditions differ across rows → partially comparable3−29.2 ptvs Grok 4.3obs. 16 Sept 2026artificialanalysis.aiT2
Humanity's Last Examknowledge · accuracy · evaluator=Artificial AnalysisIndependentevaluatorArtificial Analysisreasoningonconditions differ across rows → partially comparable3−44.3 ptvs Claude Fable 5.1obs. 16 Sept 2026artificialanalysis.aiT2
GPQA Diamondreasoning · accuracy · variant=Diamond · evaluator=Artificial AnalysisIndependentevaluatorArtificial AnalysisvariantDiamondreasoningonconditions differ across rows → partially comparable2non-primary groupobs. 16 Sept 2026artificialanalysis.aiT2
GPQA Diamondreasoning · accuracy · variant=GPQA Diamond · evaluator=Artificial AnalysisIndependentevaluatorArtificial AnalysisvariantGPQA Diamond1−16.6 ptvs gpt-6-astraobs. 11 Sept 2026artificialanalysis.aiT2

Current rows only, grouped by benchmark → canonical metric → comparability group (task configuration). Effort variants folded into this model appear as rows of the same group. 18 current rows in total. “vs leader” compares with the current leader of the benchmark's primary group only; other groups are not directly comparable. Comparability rules →

Providers & Pricing2

All offers in the price terminal →
Provider deployments, cheapest output first
ProviderContextInput / 1MOutput / 1MStatusObservedSource
DeepSeek APIcheapest outputdeepseek/deepseek-v3.2-exp163.8Kout 65.5Kactive5 d agosince 11 Sept 2026openrouter.aiT2
OpenRouterdeepseek/deepseek-v3.2-exp163.8Kout 65.5Kactive26 min agosince 12 Sept 2026openrouter.aiT2

USD per 1M tokens as published by each provider; native units (per-request fees, flex/priority tiers) are kept verbatim. Rows are append-only — every price change is kept in the history below. Cost of a workload →

Price history

Step lines per provider; amber markers are recorded changes. Click a marker or a row for the evidence behind that price.

Output price · USD / 1M tokens 2 providers

Output price history of DeepSeek V3.2 Exp$0$0.10$0.20$0.30$0.40$0.50Sept 26Sept 26Sept 26Sept 26Sept 26DeepSeek API: first observed → $0.41 · 11 Sept 2026OpenRouter: first observed → $0.41 · 12 Sept 2026
  • DeepSeek API
  • OpenRouter
  • OpenRouterfirst observed $0.4112 Sept 2026
  • DeepSeek APIfirst observed $0.4111 Sept 2026

Input price · USD / 1M tokens 2 providers

Input price history of DeepSeek V3.2 Exp$0$0.10$0.20$0.30$0.40Sept 26Sept 26Sept 26Sept 26Sept 26DeepSeek API: first observed → $0.27 · 11 Sept 2026OpenRouter: first observed → $0.27 · 12 Sept 2026
  • DeepSeek API
  • OpenRouter
  • OpenRouterfirst observed $0.2712 Sept 2026
  • DeepSeek APIfirst observed $0.2711 Sept 2026

Lineage

Open in Graph →
Explicit derived_from / fine_tuned_from / distilled_from relations stated by sources; artifacts collapsed by kind.
ANCESTORS 1V3.1-TerminusV3.1-TerminusDeepSeek V3.2 Expthis modelDeepSeek V3.2 Exp — this model

Versions & Artifacts0

Version history

Context window2 changes

11 Sept 2026→11 Sept 2026→12 Sept 2026current

Openness1 change

11 Sept 2026→11 Sept 2026

Status1 change

11 Sept 2026→11 Sept 2026

Knowledge cutofffirst observation only

11 Sept 2026current

Max outputfirst observation only

11 Sept 2026current

Each hop is a claim: click a value for its source, tier and observation time. Nothing is overwritten — a new observation closes the previous claim.

Artifacts 0

No artifact (checkpoint, quantisation, conversion or packaging) points to this model yet.

Timeline14

Full timeline →

Change history

Temporal, append-only claims: a new observation closes the previous claim instead of overwriting it. Rewind the record with the as-of picker.
1 claims · 1 propertiesShow all properties

Paperpaper_url1

Claim history for Paper
ValueValid from → toStatusSourceConfidenceExtractor
https://github.com/deepseek-ai/DeepSeek-V3.2-Exp/blob/main/DeepSeek_V3_2.pdf → currentcurrentDeepSeek — site & API docsT2mediumllm

Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →

Provenance

Attributed facts

29

Source tiers

T229

Freshest observation

5 d ago

Conflicts

None

Source documents 3

Source documents
SourceDocumentTypeTierLast observedSnapshots
DeepSeek — site & API docsapi-docs.deepseek.com/news/news250929 newsT1· Official5 d ago1
Artificial Analysisartificialanalysis.ai/leaderboards/models leaderboardT2· Quality secondary21 min ago18
OpenRouter public model & pricing listingopenrouter.ai/api/v1/models catalogueT2· Quality secondary26 min ago76

Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.

Data quality (67/100) measures how well AI Atlas knows this entity — completeness, primary-source ratio, freshness, conflicts — never how good the model is. Methodology →