DeepSeek: DeepSeek V3.2 Exp

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-gra…

deepseek/deepseek-v3.2-exp

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 designed to improve training and inference efficiency in long-context scenarios while maintaining output quality. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config)

The model was trained under conditions aligned with V3.1-Terminus to enable direct comparison. Benchmarking shows performance roughly on par with V3.1 across reasoning, coding, and agentic tool-use tasks, with minor tradeoffs and gains depending on the domain. This release focuses on validating architectural optimizations for extended context lengths rather than advancing raw task accuracy, making it primarily a research-oriented model for exploring efficient transformer designs.

Specifications & capabilities

Input
text
Output
text
Context window · tokens
163840
Knowledge cutoff
2025-07-31
Published model ID
DeepSeek-V3.2-Exp
Capabilities
reasoning, structuredOutputs, toolCalling
Description source ↗

Evaluations

61–80 / 102

LMArena

1,445.411

source id
lmarena
leaderboard name
LMArena Text Style Control French
dimension
style
metric
Arena rating
rank
116
sample size
227
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1405.9045480112418
confidence upper
1484.9168029731882
Original result ↗

LMArena

1,466.724

source id
lmarena
leaderboard name
LMArena Text Style Control Chinese
dimension
style
metric
Arena rating
rank
117
sample size
1031
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1448.8233865875573
confidence upper
1484.623814359602
Original result ↗

LMArena

1,419.442

source id
lmarena
leaderboard name
LMArena Text Factuality Russian
dimension
factuality
metric
Arena rating
rank
117
sample size
660
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1397.9652721371874
confidence upper
1440.918561764576
Original result ↗

LMArena

1,421.283

source id
lmarena
leaderboard name
LMArena Text Multi Turn
dimension
general
metric
Arena rating
rank
118
sample size
1984
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1408.312477263371
confidence upper
1434.2528286628244
Original result ↗

LMArena

1,476.86

source id
lmarena
leaderboard name
LMArena Text Factuality Coding
dimension
factuality
metric
Arena rating
rank
119
sample size
1633
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1462.8096982200134
confidence upper
1490.910942484598
Original result ↗

LMArena

1,446.85

source id
lmarena
leaderboard name
LMArena Text Style Control Hard Prompts
dimension
style
metric
Arena rating
rank
122
sample size
6557
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1438.9187389076235
confidence upper
1454.781904538503
Original result ↗

LMArena

1,425.951

source id
lmarena
leaderboard name
LMArena Text Industry Medicine And Healthcare
dimension
general
metric
Arena rating
rank
123
sample size
581
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1402.1027903319646
confidence upper
1449.8001411298671
Original result ↗

LMArena

1,413.829

source id
lmarena
leaderboard name
LMArena Text Style Control Polish
dimension
style
metric
Arena rating
rank
123
sample size
439
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1386.112121703873
confidence upper
1441.5461376498088
Original result ↗

LMArena

1,412.251

source id
lmarena
leaderboard name
LMArena Text Industry Business And Management And Financial Operations
dimension
general
metric
Arena rating
rank
123
sample size
2243
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1399.8036947884439
confidence upper
1424.6978118512632
Original result ↗

LMArena

1,404.417

source id
lmarena
leaderboard name
LMArena Text Non English
dimension
general
metric
Arena rating
rank
124
sample size
6901
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1396.6033416804212
confidence upper
1412.2304011738815
Original result ↗

LMArena

1,437.286

source id
lmarena
leaderboard name
LMArena Text Style Control English
dimension
style
metric
Arena rating
rank
125
sample size
5020
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1428.531044735133
confidence upper
1446.0400306346864
Original result ↗

LMArena

1,427.456

source id
lmarena
leaderboard name
LMArena Text Style Control Multi Turn
dimension
style
metric
Arena rating
rank
126
sample size
1984
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1414.3802253331482
confidence upper
1440.5325732408605
Original result ↗

LMArena

1,452.01

source id
lmarena
leaderboard name
LMArena Text Style Control Hard Prompts English
dimension
style
metric
Arena rating
rank
127
sample size
2753
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1440.8632754532018
confidence upper
1463.1558791102736
Original result ↗

LMArena

1,422.153

source id
lmarena
leaderboard name
LMArena Text Style Control Industry Business And Management And Financial Operations
dimension
style
metric
Arena rating
rank
127
sample size
2243
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1409.6872812216334
confidence upper
1434.6187292102027
Original result ↗

LMArena

1,429.614

source id
lmarena
leaderboard name
LMArena Text Industry Life And Physical And Social Science
dimension
general
metric
Arena rating
rank
128
sample size
1930
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1416.289874974726
confidence upper
1442.9375194394406
Original result ↗

LMArena

1,414.882

source id
lmarena
leaderboard name
LMArena Text Style Control Exclude Ties
dimension
style
metric
Arena rating
rank
128
sample size
8396
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1405.9704654300158
confidence upper
1423.7927451621424
Original result ↗

LMArena

1,414.422

source id
lmarena
leaderboard name
LMArena Text Style Control Instruction Following
dimension
style
metric
Arena rating
rank
129
sample size
3378
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1404.18447406607
confidence upper
1424.6598357280818
Original result ↗

LMArena

1,417.484

source id
lmarena
leaderboard name
LMArena Text Math
dimension
general
metric
Arena rating
rank
130
sample size
783
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1397.4084268243394
confidence upper
1437.5590780412122
Original result ↗

LMArena

1,401.189

source id
lmarena
leaderboard name
LMArena Text Instruction Following
dimension
general
metric
Arena rating
rank
130
sample size
3378
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1390.965673426147
confidence upper
1411.412654087961
Original result ↗

LMArena

1,403.541

source id
lmarena
leaderboard name
LMArena Text Russian
dimension
general
metric
Arena rating
rank
131
sample size
1170
evaluated at
2026-11-09T00:00:00.000Z
source model id
DeepSeek-V3.2-Exp
confidence lower
1387.0491631785005
confidence upper
1420.0331386088653
Original result ↗

Published prices

1–6 / 6

deepseek-ai/DeepSeek-V3.2-Exp ↗

$0.27

Inference provider
siliconFlow
Platform
SiliconFlow
Unit
input_tokens_per_million
Currency
USD
Billing conditions
Billing label
Input tokens (per 1M)

deepseek-ai/DeepSeek-V3.2-Exp ↗

$0.41

Inference provider
siliconFlow
Platform
SiliconFlow
Unit
output_tokens_per_million
Currency
USD
Billing conditions
Billing label
Output tokens (per 1M)

Provider specifications & access

1–3 / 3

DeepSeek-V3.2-Exp ↗

source model id
deepseek-ai/DeepSeek-V3.2-Exp
hosting
siliconflow-hosted
platforms
  • SiliconFlow
max output tokens
163840
architecture
input_modalities
  • text
output_modalities
  • text
capabilities
  • reasoning
  • structuredOutputs
  • toolCalling

deepseek-ai/DeepSeek-V3.2-Exp ↗

source model id
publishers/deepseek-ai/models/deepseek-v3-2@deepseek-v3-2-exp
hosting
vertex-self-deployed
platforms
  • Google Cloud Vertex AI
architecture
input_modalities
  • text
output_modalities
  • text
capabilities

    deepseek-ai/DeepSeek-V3.2-Exp ↗

    source model id
    publishers/hf-deepseek-ai/models/deepseek-v3.2-exp@001
    hosting
    vertex-self-deployed
    platforms
    • Google Cloud Vertex AI
    architecture
    input_modalities
    • text
    output_modalities
    • text
    capabilities

      Safety assessments

      1–1 / 1

      Foundation Model Transparency Index 2025 ↗

      32 / 100

      source id
      stanfordFMTI
      canonical model id
      deepseek/deepseek-v3.2-exp
      taxonomy
      transparency
      hazard category
      transparency
      hazard name
      Organization transparency
      policy area
      DeepSeek
      assessment type
      organization-level disclosure
      metric
      disclosure_score
      result label
      32 / 100
      sample size
      100
      updated at
      2026-11-09T18:08:05.227Z

      Source specification observations

      1–5 / 5

      openRouter ↗

      field
      context_length
      value
      163840
      normalized value
      163840
      scope
      source-model
      source id
      openRouter
      confidence
      1

      siliconFlow ↗

      field
      context_length
      value
      164K
      scope
      source-model
      source id
      siliconFlow
      confidence
      1

      openRouter ↗

      field
      knowledge_cutoff
      reported field
      knowledge_cutoff
      value
      2025-07-31
      normalized value
      2025-07-31
      precision
      day
      scope
      source-model
      source id
      openRouter
      confidence
      1

      openRouter ↗

      field
      max_output_tokens
      value
      147456
      normalized value
      147456
      scope
      source-model
      source id
      openRouter
      confidence
      1

      siliconFlow ↗

      field
      max_output_tokens
      value
      164K
      scope
      source-model
      source id
      siliconFlow
      confidence
      1

      Original sources

      1–19 / 19

      DeepSeek · Model reference ↗

      SiliconFlow · Model page ↗

      Hugging Face · Model repository ↗

      DeepInfra · Model page ↗

      Google Cloud · Model page ↗

      Google Cloud · Model page ↗

      OpenRouter · Model page ↗

      SiliconFlow · Playground ↗

      SiliconFlow · API documentation ↗

      GitHub · Model reference ↗

      DeepInfra · Catalog ↗

      Epoch AI · Catalog ↗

      GitHub · Catalog ↗

      GitHub · Catalog ↗

      Google Cloud · Catalog ↗

      LMArena · Catalog ↗

      OpenRouter · Catalog ↗

      SiliconFlow · Catalog ↗

      gorilla.cs.berkeley.edu · Catalog ↗

      Source links

      1–11 / 11

      Explore more models

      Models

      Loading
      EVIDENCE / COMPARISON

      Benchmarks

      Explore every benchmark

      Loading

      These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

      Install TensorTrace