Model comparison
DeepSeek-V3.2-Speciale vs Llama 13b
DeepSeek-V3.2-Speciale is the stronger model overall, scoring 39.1 to 23.9 on the Noometry Index.
Last verified . 0 shared benchmarks.
DeepSeek
MetaSummary
- The widest gap is in writing & preference, where DeepSeek-V3.2-Speciale leads 46.1 to 13.9.
Side by side
| DeepSeek-V3.2-Speciale | Llama 13b | |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 39.1 | 23.9 |
| Released | 2025-12-01 | 2023-02-24 |
| Weights | Open | Open |
| Context window | 128K | — |
| Max output | 128K | — |
| Input $ / M tokens | $0.58 | — |
| Output $ / M tokens | $1.68 | — |
| Results tracked | 3 | 21 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.2-Speciale leads
DeepSeek-V3.2-Speciale: 39.9 (#136), Llama 13b: 20.9 (#332)
| Benchmark | DeepSeek-V3.2-Speciale | Llama 13b |
|---|---|---|
| WeirdML | 46.7% | — |
| LMArena Coding | — | 683 |
Reasoning DeepSeek-V3.2-Speciale leads
DeepSeek-V3.2-Speciale: 32.1 (#65), Llama 13b: 13.4 (#327)
| Benchmark | DeepSeek-V3.2-Speciale | Llama 13b |
|---|---|---|
| SimpleBench | 52.6% | — |
| LMArena Hard Prompts | — | 728 |
| BIG-Bench Hard | — | 37.9% |
| Epoch Capabilities Index | — | 100.09 |
| HellaSwag | — | 79.2% |
| LAMBADA | — | 75.2% |
| PIQA | — | 80.1% |
| WinoGrande | — | 73% |
Math Not comparable
DeepSeek-V3.2-Speciale: —, Llama 13b: 24.4 (#259)
| Benchmark | DeepSeek-V3.2-Speciale | Llama 13b |
|---|---|---|
| LMArena Math | — | 838 |
| GSM8K | — | 20.6% |
Knowledge Not comparable
DeepSeek-V3.2-Speciale: —, Llama 13b: —
| Benchmark | DeepSeek-V3.2-Speciale | Llama 13b |
|---|---|---|
| ARC (AI2) Challenge | — | 52.7% |
| BoolQ | — | 78.7% |
| MMLU | — | 47.7% |
| OpenBookQA | — | 56.4% |
| TriviaQA | — | 77.9% |
Multimodal Not comparable
DeepSeek-V3.2-Speciale: —, Llama 13b: —
| Benchmark | DeepSeek-V3.2-Speciale | Llama 13b |
|---|---|---|
| ScienceQA | — | 43.3% |
Multilingual Not comparable
DeepSeek-V3.2-Speciale: —, Llama 13b: 16.8 (#292)
| Benchmark | DeepSeek-V3.2-Speciale | Llama 13b |
|---|---|---|
| LMArena Non-English | — | 819 |
Instruction Following Not comparable
DeepSeek-V3.2-Speciale: —, Llama 13b: 36.9 (#300)
| Benchmark | DeepSeek-V3.2-Speciale | Llama 13b |
|---|---|---|
| LMArena Instruction Following | — | 781 |
Writing & Preference DeepSeek-V3.2-Speciale leads
DeepSeek-V3.2-Speciale: 46.1 (#217), Llama 13b: 13.9 (#307)
| Benchmark | DeepSeek-V3.2-Speciale | Llama 13b |
|---|---|---|
| LMArena Text | — | 834 |
| LMArena Creative Writing | — | 794 |
| EQ-Bench Creative Writing | 1276 | — |
| LMArena Multi-Turn | — | 753 |
Frequently asked questions
Is DeepSeek-V3.2-Speciale better than Llama 13b?
DeepSeek-V3.2-Speciale is the stronger model overall, scoring 39.1 to 23.9 on the Noometry Index.
Is DeepSeek-V3.2-Speciale or Llama 13b better for coding?
DeepSeek-V3.2-Speciale scores higher on coding benchmarks: 39.9 versus 20.9 in the Noometry coding category.
How many benchmarks do DeepSeek-V3.2-Speciale and Llama 13b share?
0 benchmarks have published results for both models. DeepSeek-V3.2-Speciale has 3 scored results on Noometry and Llama 13b has 21.