Model profile
DeepSeek V3.2 (Thinking)
Evidence coverage
2 of 321 tracked benchmarks are published. 1 is verified and 1 provisional. 2 of 8 categories are measured.
- Published / tracked
- 2 / 321
- Verified
- 1
- Provisional
- 1
- Categories with evidence
- 2 / 8
Evidence by category
- Agentic0 benchmarksNot measured
- Coding1 benchmarkVerified
- Reasoning0 benchmarksNot measured
- Knowledge0 benchmarksNot measured
- Math0 benchmarksNot measured
- Multilingual0 benchmarksNot measured
- Multimodal1 benchmarkReported
- Inst. Following0 benchmarksNot measured
DeepSeek V3.2 (Thinking) ranks #65 out of 200 models on the public leaderboard with an overall score of 58.15/100. It does not yet have enough sourced coverage for BenchLM's verified leaderboard. While not a frontier model, it offers specific advantages depending on the use case.
DeepSeek V3.2 (Thinking) is a open weight model with a 128K token context window. It uses explicit chain-of-thought reasoning, which typically improves performance on math and complex reasoning tasks at the cost of higher latency and token usage.
DeepSeek V3.2 (Thinking) sits inside the DeepSeek V3.2 family alongside DeepSeek V3.2. This profile currently has 2 of 321 tracked benchmarks. BenchLM only exposes non-generated benchmark rows publicly, so missing categories stay blank until a sourced evaluation is available.
Its strongest category is Coding (#56). This performance profile makes it particularly well-suited for software development and code generation tasks.
Peer position
Exact provisional scores and ranks for the closest listed peers. A score can appear before a model clears the evidence threshold for a rank, so equal scores can have different rank states.
Range 57.56–58.62
- MiMo-V2.5XiaomiCompare#6258.62MiMo-V2.5 is #62 with a score of 58.62.
- GPT-5 (high)OpenAICompare#6358.61GPT-5 (high) is #63 with a score of 58.61.
- GPT-5.2OpenAICompare#6458.43GPT-5.2 is #64 with a score of 58.43.
- DeepSeek V3.2 (Thinking)Current modelDeepSeek#6558.15DeepSeek V3.2 (Thinking) is #65 with a score of 58.15.
- Qwen3 235B 2507 (Reasoning)AlibabaCompare#6658.01Qwen3 235B 2507 (Reasoning) is #66 with a score of 58.01.
- Gemma 4 26B A4BGoogleCompare#6757.96Gemma 4 26B A4B is #67 with a score of 57.96.
- GLM-4.5Z.AICompare#6857.56GLM-4.5 is #68 with a score of 57.56.
Category percentile
More
Relative position among models eligible for each sourced category. A higher percentile means a stronger position within that category's ranked cohort; 100 is highest.
- Coding55%Eligible cohort rank #56 of 122Category score 51.8
Category evidence
Scores and ranks appear only where this model has published benchmark evidence. Categories without displayable source records remain not measured.
| Category | Score | Rank | Percentile | Weight | Benchmarks | Evidence |
|---|---|---|---|---|---|---|
| AgenticWeight 22%0 benchmarksNot measured | Not measured | Not ranked | Not available | 22% | 0 benchmarks | Not measured |
| CodingRank #56 of 122Percentile 55thWeight 20%1 benchmarkVerified | 51.8 | #56 of 122 | 55th | 20% | 1 benchmark | Verified |
| ReasoningWeight 17%0 benchmarksNot measured | Not measured | Not ranked | Not available | 17% | 0 benchmarks | Not measured |
| KnowledgeWeight 12%0 benchmarksNot measured | Not measured | Not ranked | Not available | 12% | 0 benchmarks | Not measured |
| MathWeight 5%0 benchmarksNot measured | Not measured | Not ranked | Not available | 5% | 0 benchmarks | Not measured |
| MultilingualWeight 7%0 benchmarksNot measured | Not measured | Not ranked | Not available | 7% | 0 benchmarks | Not measured |
| MultimodalRank Not rankedWeight 12%1 benchmarkReported | 70.4 | Not ranked | Not available | 12% | 1 benchmark | Reported |
| Inst. FollowingWeight 5%0 benchmarksNot measured | Not measured | Not ranked | Not available | 5% | 0 benchmarks | Not measured |
Chatbot Arena performance
Scroll horizontally to inspect confidence intervals and vote counts.
| View | Elo | Confidence interval | Votes |
|---|---|---|---|
| Text Overall | 1423 | ±3.6 | 41,030 |
| Coding | 1475 | ±6.8 | 8,535 |
| Math | 1425 | ±11.8 | 2,495 |
| Instruction Following | 1419 | ±6.1 | 11,146 |
| Creative Writing | 1391 | ±7.9 | 6,193 |
| Multi-turn | 1427 | ±7.6 | 6,874 |
| Hard Prompts | 1446 | ±4.7 | 22,451 |
| Hard Prompts (English) | 1462 | ±6.3 | 10,592 |
| Longer Query | 1442 | ±6.2 | 10,971 |
Benchmark Details
Rows below have a displayable published verification record. Each source link and provenance note remains in the page HTML while its category is closed. Source-unverified manual rows and generated rows stay hidden.
Coding1 benchmark
Vibe Code Bench v1.1
Multimodal1 benchmark
Design Arena Website Elo
Frequently Asked Questions
How does DeepSeek V3.2 (Thinking) perform overall in AI benchmarks?
DeepSeek V3.2 (Thinking) has 2 published benchmark scores on BenchLM, but it does not yet have enough non-generated coverage to receive a global overall rank.
Is DeepSeek V3.2 (Thinking) good for coding and programming?
DeepSeek V3.2 (Thinking) ranks #56 out of 122 models in coding and programming benchmarks with an average score of 51.8. There are stronger options in this category.
Is DeepSeek V3.2 (Thinking) good for multimodal and grounded tasks?
DeepSeek V3.2 (Thinking) has visible benchmark coverage in multimodal and grounded tasks, but BenchLM does not currently assign it a global category rank there.
Is DeepSeek V3.2 (Thinking) open source?
Yes, DeepSeek V3.2 (Thinking) is an open weight model created by DeepSeek, meaning it can be downloaded and run locally or fine-tuned for specific use cases.
Which sibling models are related to DeepSeek V3.2 (Thinking)?
DeepSeek V3.2 (Thinking) belongs to the DeepSeek V3.2 family. Related variants on BenchLM include DeepSeek V3.2.
Does DeepSeek V3.2 (Thinking) have full benchmark coverage on BenchLM?
Not yet. DeepSeek V3.2 (Thinking) currently has 2 published benchmark scores out of the 321 benchmarks BenchLM tracks. BenchLM only exposes non-generated public benchmark rows, so missing categories stay blank until a sourced evaluation is available.
What is the context window size of DeepSeek V3.2 (Thinking)?
DeepSeek V3.2 (Thinking) has a published context window of 128K, which determines how much text it can process in a single interaction.
Related Resources
Choose with this week’s evidence
Join 2,000+ readers for ranking moves, new releases, pricing changes, and the evidence behind them.
Free. One email per week.