# DeepSeek V3.2 vs DeepSeek V3.2 (Thinking)

> Side-by-side benchmark comparison for DeepSeek V3.2 and DeepSeek V3.2 (Thinking) across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/deepseek-v3-2-vs-deepseek-v3-2-thinking
- Last updated: September 4, 2026

- Shared sourced benchmarks: 0
- HTML indexing: indexable
- Ranking lane: BenchAlign v5

## Quick Verdict

No benchmark verdict yet: there is no shared sourced benchmark row for DeepSeek V3.2 and DeepSeek V3.2 (Thinking).

## Summary

- DeepSeek V3.2 and DeepSeek V3.2 (Thinking) do not currently share a sourced benchmark row. The mirror therefore compares metadata, pricing, context windows, and each model's separately reported coverage without naming a benchmark winner.

## Model Snapshot

| Property | DeepSeek V3.2 | DeepSeek V3.2 (Thinking) |
|----------|----------|----------|
| Creator | DeepSeek | DeepSeek |
| Type | Open Weight | Open Weight |
| Reasoning | Non-Reasoning | Reasoning |
| Context | 128K | 128K |
| Independent public score (not head-to-head) | 57.62 | 57.54 |
| Benchmarks Covered | 7 | 1 |
| Pricing (input/output) | $0.28 / $0.42 | $0.55 / $2.19 |

## Category Breakdown

### Agentic

- Winner: Insufficient shared evidence
- DeepSeek V3.2 public-lane score: Coming soon
- DeepSeek V3.2 (Thinking) public-lane score: Coming soon

| Benchmark | DeepSeek V3.2 | DeepSeek V3.2 (Thinking) | Winner |
|-----------|-----------|-----------|--------|
| Claw-Eval | 40.2% | Coming soon | Coming soon |
| VITA-Bench | 18.5% | Coming soon | Coming soon |
| Gert Labs | 29.57% | Coming soon | Coming soon |

### Coding

- Winner: Insufficient shared evidence
- DeepSeek V3.2 public-lane score: 37.2 (Supported · #152/183)
- DeepSeek V3.2 (Thinking) public-lane score: 49.6 (Estimated · #76/183)

| Benchmark | DeepSeek V3.2 | DeepSeek V3.2 (Thinking) | Winner |
|-----------|-----------|-----------|--------|
| SWE-Rebench | 60.9% | Coming soon | Coming soon |
| React Native Evals | 71.5% | Coming soon | Coming soon |
| Vibe Code Bench | Coming soon | 5.11% | Coming soon |

### Reasoning

- Winner: Insufficient shared evidence
- DeepSeek V3.2 public-lane score: 51.7 (Unranked · 2 rankable rows)
- DeepSeek V3.2 (Thinking) public-lane score: Coming soon

### Knowledge

- Winner: Insufficient shared evidence
- DeepSeek V3.2 public-lane score: 49.7 (Estimated · #92/181)
- DeepSeek V3.2 (Thinking) public-lane score: Coming soon

### Instruction Following

- Winner: Insufficient shared evidence
- DeepSeek V3.2 public-lane score: 58 (#71/120)
- DeepSeek V3.2 (Thinking) public-lane score: Coming soon

### Mathematics

- Winner: Insufficient shared evidence
- DeepSeek V3.2 public-lane score: 40.2 (Unranked · 2 rankable rows)
- DeepSeek V3.2 (Thinking) public-lane score: Coming soon

| Benchmark | DeepSeek V3.2 | DeepSeek V3.2 (Thinking) | Winner |
|-----------|-----------|-----------|--------|
| FrontierMath v2 (Tiers 1-3) | 22.100% | Coming soon | Coming soon |
| FrontierMath v2 (Tier 4) | 2.100% | Coming soon | Coming soon |

## FAQ

### Can I compare DeepSeek V3.2 and DeepSeek V3.2 (Thinking) on BenchLM yet?

Not fully yet. BenchLM is tracking both models, but sourced benchmark coverage is still incomplete for a fair score-level comparison.

### Why does this page show "coming soon" values?

BenchLM only calls winners when public benchmark coverage is available on both sides of the comparison.

## Related Comparisons

- [DeepSeek V3.2 vs Claude Fable 5.1](/compare/claude-fable-5-1-vs-deepseek-v3-2)
- [DeepSeek V3.2 (Thinking) vs Claude Fable 5.1](/compare/claude-fable-5-1-vs-deepseek-v3-2-thinking)
- [DeepSeek V3.2 vs GPT-6 Astra](/compare/deepseek-v3-2-vs-gpt-6-astra)
- [DeepSeek V3.2 (Thinking) vs GPT-6 Astra](/compare/deepseek-v3-2-thinking-vs-gpt-6-astra)

## Explore More

- [DeepSeek V3.2 profile](/models/deepseek-v3-2)
- [DeepSeek V3.2 (Thinking) profile](/models/deepseek-v3-2-thinking)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)