Model comparison
DeepSeek-V3.2-Exp vs DeepSeek-V3.2-Speciale
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.9 to 39.1 on the Noometry Index.
Last verified . 2 shared benchmarks.
DeepSeekSummary
- They share 2 benchmarks with published results for both. DeepSeek-V3.2-Exp scores higher in 2 categories and DeepSeek-V3.2-Speciale in 1 category; 3 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-V3.2-Exp leads 62.5 to 46.1.
- The biggest single-benchmark swing is WeirdML: 39.5% for DeepSeek-V3.2-Exp and 46.7% for DeepSeek-V3.2-Speciale.
- DeepSeek-V3.2-Exp is cheaper at $0.26 / $0.38 per million input/output tokens, against $0.58 / $1.68 for DeepSeek-V3.2-Speciale.
- DeepSeek-V3.2-Exp accepts more context: 164K tokens versus 128K.
Side by side
| DeepSeek-V3.2-Exp | DeepSeek-V3.2-Speciale | |
|---|---|---|
| Provider | DeepSeek | DeepSeek |
| Noometry Index | 44.9 | 39.1 |
| Released | 2025-09-29 | 2025-12-01 |
| Weights | Open | Open |
| Context window | 164K | 128K |
| Max output | 66K | 128K |
| Input $ / M tokens | $0.26 | $0.58 |
| Output $ / M tokens | $0.38 | $1.68 |
| Results tracked | 48 | 3 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 46.4 (#61), DeepSeek-V3.2-Speciale: 39.9 (#136)
| Benchmark | DeepSeek-V3.2-Exp | DeepSeek-V3.2-Speciale |
|---|---|---|
| WeirdML | 39.5% | 46.7% |
| SWE-bench Verified (bash only) | 70% | — |
| Aider Polyglot | 74.2% | — |
| LMArena WebDev | 1362 | — |
| SWE-bench Multilingual | 59% | — |
| SciCode | 38.9% | — |
| LMArena Coding | 1454 | — |
Agentic & Tool Use Not comparable
DeepSeek-V3.2-Exp: 37.9 (#31), DeepSeek-V3.2-Speciale: —
| Benchmark | DeepSeek-V3.2-Exp | DeepSeek-V3.2-Speciale |
|---|---|---|
| Terminal-Bench | 39.6% | — |
| Berkeley Function Calling Leaderboard | 56.7% | — |
| TheAgentCompany | 42.9% | — |
| Vending-Bench 2 | 1,034 | — |
Reasoning DeepSeek-V3.2-Speciale leads
DeepSeek-V3.2-Exp: 22.0 (#185), DeepSeek-V3.2-Speciale: 32.1 (#65)
| Benchmark | DeepSeek-V3.2-Exp | DeepSeek-V3.2-Speciale |
|---|---|---|
| ARC-AGI-2 | 4% | — |
| SimpleBench | — | 52.6% |
| Kagi LLM Benchmark | 52.2% | — |
| NYT Connections (extended) | 36.7% | — |
| ARC-AGI-1 | 57% | — |
| CritPt | 2.9% | — |
| Chess Puzzles | 14% | — |
| Thematic Generalization | 65% | — |
| LMArena Hard Prompts | 1433 | — |
| DTBench | 87.7% | — |
| LMCA | 29.1% | — |
| Epoch Capabilities Index | 146.28 | — |
Math Not comparable
DeepSeek-V3.2-Exp: 40.8 (#75), DeepSeek-V3.2-Speciale: —
| Benchmark | DeepSeek-V3.2-Exp | DeepSeek-V3.2-Speciale |
|---|---|---|
| MathArena Final-Answer Competitions | 57.7% | — |
| OTIS Mock AIME 2024-2025 | 87.8% | — |
| ProofBench | 8% | — |
| LMArena Math | 1435 | — |
| FrontierMath (Feb 2025 set) | 22.1% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge Not comparable
DeepSeek-V3.2-Exp: 51.6 (#61), DeepSeek-V3.2-Speciale: —
| Benchmark | DeepSeek-V3.2-Exp | DeepSeek-V3.2-Speciale |
|---|---|---|
| GPQA Diamond | 83.4% | — |
| Vectara Hallucination Rate | 5.3% | — |
| LMArena Expert | 1436 | — |
Multilingual Not comparable
DeepSeek-V3.2-Exp: 52.4 (#85), DeepSeek-V3.2-Speciale: —
| Benchmark | DeepSeek-V3.2-Exp | DeepSeek-V3.2-Speciale |
|---|---|---|
| LMArena Non-English | 1409 | — |
| LMArena Chinese | 1461 | — |
| LMArena French | 1435 | — |
| LMArena German | 1440 | — |
| LMArena Japanese | 1377 | — |
| LMArena Korean | 1372 | — |
| LMArena Russian | 1423 | — |
| LMArena Spanish | 1441 | — |
Instruction Following Not comparable
DeepSeek-V3.2-Exp: 74.7 (#89), DeepSeek-V3.2-Speciale: —
| Benchmark | DeepSeek-V3.2-Exp | DeepSeek-V3.2-Speciale |
|---|---|---|
| LMArena Instruction Following | 1413 | — |
Long Context Not comparable
DeepSeek-V3.2-Exp: 47.6 (#15), DeepSeek-V3.2-Speciale: —
| Benchmark | DeepSeek-V3.2-Exp | DeepSeek-V3.2-Speciale |
|---|---|---|
| Fiction.LiveBench | 83.3% | — |
| CL-bench | 13.2% | — |
| CL-bench Life | 9.5% | — |
| LMArena Longer Query | 1428 | — |
Writing & Preference DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 62.5 (#73), DeepSeek-V3.2-Speciale: 46.1 (#217)
| Benchmark | DeepSeek-V3.2-Exp | DeepSeek-V3.2-Speciale |
|---|---|---|
| EQ-Bench Creative Writing | 1515 | 1276 |
| LMArena Text | 1425 | — |
| LMArena Creative Writing | 1403 | — |
| LMArena Multi-Turn | 1427 | — |
Frequently asked questions
Is DeepSeek-V3.2-Exp better than DeepSeek-V3.2-Speciale?
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.9 to 39.1 on the Noometry Index.
Which is cheaper, DeepSeek-V3.2-Exp or DeepSeek-V3.2-Speciale?
DeepSeek-V3.2-Exp is cheaper. It lists at $0.26 per million input tokens and $0.38 per million output tokens; DeepSeek-V3.2-Speciale lists at $0.58 and $1.68.
Is DeepSeek-V3.2-Exp or DeepSeek-V3.2-Speciale better for coding?
DeepSeek-V3.2-Exp scores higher on coding benchmarks: 46.4 versus 39.9 in the Noometry coding category.
Which has the bigger context window?
DeepSeek-V3.2-Exp does, with 164K tokens against 128K.
How many benchmarks do DeepSeek-V3.2-Exp and DeepSeek-V3.2-Speciale share?
2 benchmarks have published results for both models. DeepSeek-V3.2-Exp has 48 scored results on Noometry and DeepSeek-V3.2-Speciale has 3.