DeepSeek V3.2 Thinking
DeepSeek V3.2 (thinking/reasoner mode) — official successor to V3.2-Exp. Reasoning-first model built for agents with GPT-5 level performance. Balanced inference vs. output length. First DeepSeek model with thinking-in-tool-use capability. FP8.
- Reasoning
- Native PDF input
- Tool Calling
- Structured Output
Added Dec 1, 2025
Model weightsPricing
Auto routing · per 1M tokens- Input
- $0.26
- Output
- $0.38
- Cache read
- $0.13
Specifications
- Context window
- 163K
- Max output
- 65.5K
- Parameters
- 685B / 37B
- Total / active
Benchmarks
Benchmarks
Sourced from LMArena.
Arena Score
1422.7
Overall Rank
#129 / 421
Votes
41,763
Confidence Interval
1419.1 - 1426.2
Category Scores
Coding
#121 / 416
8,826 votes
1475.4
Math
#112 / 404
2,610 votes
1428.5
Longer Query
#112 / 399
11,635 votes
1442.1
Creative Writing
#127 / 419
6,346 votes
1391.1
Instruction Following
#114 / 421
11,641 votes
1419.8
Hard Prompts
#123 / 421
23,172 votes
1446.3
Additional Categories22
Industry Mathematical
#96 / 402
1,934 votes
1449.5
Japanese
#106 / 288
382 votes
1386.8
Hard Prompts English
#109 / 419
9,526 votes
1460.5
English
#112 / 421
17,031 votes
1442.5
Expert
#115 / 371
2,860 votes
1453.8
Polish
#117 / 243
1,115 votes
1418.4
German
#119 / 320
741 votes
1418.5
Korean
#119 / 289
697 votes
1371.8
Industry Software And It Services
#121 / 421
14,192 votes
1464.3
Industry Business And Management And Financial Operations
#124 / 414
8,065 votes
1422.8
Multi Turn
#125 / 419
6,969 votes
1427.6
Industry Writing And Literature And Language
#126 / 420
9,646 votes
1403.7
Spanish
#130 / 304
1,356 votes
1413.7
Industry Legal And Government
#132 / 393
3,068 votes
1429.5
Exclude Ties
#133 / 421
29,252 votes
1412.2
Industry Life And Physical And Social Science
#134 / 419
6,815 votes
1438.2
Chinese
#135 / 395
3,089 votes
1454.0
Industry Entertainment And Sports And Media
#135 / 419
7,691 votes
1388.2
French
#136 / 301
1,038 votes
1429.5
Russian
#140 / 385
5,393 votes
1407.0
Non English
#141 / 421
24,712 votes
1401.1
Industry Medicine And Healthcare
#148 / 391
2,805 votes
1432.0
Published 2026-11-11 · Matched as DeepSeek-V3.2 (Thinking) · CC BY 4.0
LMArena DatasetLeaderboard results
These leaderboards tested DeepSeek V3.2, another version of this model.
| Benchmark | Score | Rank on NanoGPT |
|---|---|---|
Epoch AI Benchmarking Hub Epoch AI data, CC BY 4.0 · Updated Nov 13, 2026 | ||
Epoch Capabilities Index | 146.3 | #79 of 115 |
Creative Writing v3 | 1515 | #55 of 96 |
Longform Writing | 66.8 | #46 of 97 |
UGI Leaderboard UGI Leaderboard data, Apache-2.0 · Updated Oct 16, 2026 | ||
UGI Score | 57.8 | #13 of 149 |
UGI Willingness | 7.2 | #9 of 149 |
UGI Writing | 57.3 | #43 of 149 |
Filtered and ranked to models on NanoGPT.
Providers
Choose or restrict providers with the API. Provider-selection prices include the selection markup. Provider selection guide · Pricing API (JSON, USD per 1K tokens)
 | $0.26 | $0.38 | $0.13 | N/A | N/A |
| $0.26 | $0.38 | $0.13 | N/A | N/A |
| $0.305$0.219 | $0.452$0.325 | $0.0315$0.0227 | N/A | N/A | |
| $0.272 | $0.84 | $0.147 | N/A | N/A | |
| $0.273 | $0.399 | $0.136 | N/A | N/A | |
| $0.273 | $0.399 | $0.136 | N/A | N/A | |
| $0.278 | $0.526 | N/A | N/A | N/A | |
| $0.294 | $0.441 | $0.0294 | N/A | N/A |
Batch rows for this model use normal routing and cost the same as standard API requests. Set provider in a row to choose a provider.