Model comparison
DeepSeek-V3.2-Exp vs Llama 2-13B
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 29.6 on the Noometry Index.
Last verified . 20 shared benchmarks.
DeepSeek
MetaSummary
- They share 20 benchmarks with published results for both. DeepSeek-V3.2-Exp scores higher in 8 categories and Llama 2-13B in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-V3.2-Exp leads 62.4 to 29.8.
- The biggest single-benchmark swing is DTBench: 87.7% for DeepSeek-V3.2-Exp and 42.2% for Llama 2-13B.
Side by side
| DeepSeek-V3.2-Exp | Llama 2-13B | |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 44.3 | 29.6 |
| Released | 2025-09-29 | 2023-07-18 |
| Weights | Open | Open |
| Context window | 164K | — |
| Max output | 66K | — |
| Input $ / M tokens | $0.26 | — |
| Output $ / M tokens | $0.38 | — |
| Results tracked | 49 | 32 |
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: 46.5 (#65), Llama 2-13B: 30.9 (#291)
| Benchmark | DeepSeek-V3.2-Exp | Llama 2-13B |
|---|---|---|
| LMArena Coding | 1454 | 1062 |
| SWE-bench Verified (bash only) | 70% | — |
| Aider Polyglot | 74.2% | — |
| LMArena WebDev | 1362 | — |
| SWE-bench Multilingual | 59% | — |
| SciCode | 38.9% | — |
| WeirdML | 39.5% | — |
Agentic & Tool Use Not comparable
DeepSeek-V3.2-Exp: 32.7 (#59), Llama 2-13B: —
| Benchmark | DeepSeek-V3.2-Exp | Llama 2-13B |
|---|---|---|
| Terminal-Bench | 39.6% | — |
| APEX-Agents | 21.3% | — |
| Berkeley Function Calling Leaderboard | 56.7% | — |
| TheAgentCompany | 42.9% | — |
| Vending-Bench 2 | 1,034 | — |
Reasoning DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 22.1 (#208), Llama 2-13B: 12.8 (#337)
| Benchmark | DeepSeek-V3.2-Exp | Llama 2-13B |
|---|---|---|
| Chess Puzzles | 14% | 0% |
| LMArena Hard Prompts | 1434 | 1051 |
| DTBench | 87.7% | 42.2% |
| Epoch Capabilities Index | 146.27 | 106.17 |
| ARC-AGI-2 | 4% | — |
| Kagi LLM Benchmark | 52.2% | — |
| NYT Connections (extended) | 36.7% | — |
| ARC-AGI-1 | 57% | — |
| CritPt | 2.9% | — |
| Thematic Generalization | 65% | — |
| LMCA | 29.1% | — |
| BIG-Bench Hard | — | 58.2% |
| HellaSwag | — | 80.7% |
| LAMBADA | — | 76.5% |
| PIQA | — | 80.8% |
| WinoGrande | — | 72.8% |
Math DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 41.7 (#87), Llama 2-13B: 31.1 (#229)
| Benchmark | DeepSeek-V3.2-Exp | Llama 2-13B |
|---|---|---|
| LMArena Math | 1435 | 1065 |
| MathArena Final-Answer Competitions | 57.7% | — |
| OTIS Mock AIME 2024-2025 | 87.8% | — |
| ProofBench | 8% | — |
| FrontierMath (Feb 2025 set) | 22.1% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
| GSM8K | — | 36.9% |
Knowledge DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 51.7 (#66), Llama 2-13B: 28.1 (#249)
| Benchmark | DeepSeek-V3.2-Exp | Llama 2-13B |
|---|---|---|
| LMArena Expert | 1436 | 1030 |
| GPQA Diamond | 83.4% | — |
| Vectara Hallucination Rate | 5.3% | — |
| ARC (AI2) Challenge | — | 60.3% |
| BoolQ | — | 82.4% |
| MMLU | — | 55.6% |
| OpenBookQA | — | 57% |
| TriviaQA | — | 79.6% |
Multimodal Not comparable
DeepSeek-V3.2-Exp: —, Llama 2-13B: —
| Benchmark | DeepSeek-V3.2-Exp | Llama 2-13B |
|---|---|---|
| ScienceQA | — | 55.8% |
Multilingual DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 52.2 (#90), Llama 2-13B: 26.5 (#279)
| Benchmark | DeepSeek-V3.2-Exp | Llama 2-13B |
|---|---|---|
| LMArena Non-English | 1409 | 1024 |
| LMArena Chinese | 1461 | 1001 |
| LMArena French | 1433 | 1044 |
| LMArena German | 1440 | 1009 |
| LMArena Japanese | 1374 | 894 |
| LMArena Korean | 1371 | 953 |
| LMArena Russian | 1424 | 1055 |
| LMArena Spanish | 1440 | 1087 |
Instruction Following DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 74.5 (#93), Llama 2-13B: 53.3 (#287)
| Benchmark | DeepSeek-V3.2-Exp | Llama 2-13B |
|---|---|---|
| LMArena Instruction Following | 1413 | 1045 |
Long Context DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 47.6 (#16), Llama 2-13B: 32.3 (#269)
| Benchmark | DeepSeek-V3.2-Exp | Llama 2-13B |
|---|---|---|
| LMArena Longer Query | 1428 | 1064 |
| Fiction.LiveBench | 83.3% | — |
| CL-bench | 13.2% | — |
| CL-bench Life | 9.5% | — |
Writing & Preference DeepSeek-V3.2-Exp leads
DeepSeek-V3.2-Exp: 62.4 (#77), Llama 2-13B: 29.8 (#289)
| Benchmark | DeepSeek-V3.2-Exp | Llama 2-13B |
|---|---|---|
| LMArena Text | 1425 | 1084 |
| LMArena Creative Writing | 1403 | 1047 |
| LMArena Multi-Turn | 1427 | 1050 |
| EQ-Bench Creative Writing | 1515 | — |
Frequently asked questions
Is DeepSeek-V3.2-Exp better than Llama 2-13B?
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 29.6 on the Noometry Index.
Is DeepSeek-V3.2-Exp or Llama 2-13B better for coding?
DeepSeek-V3.2-Exp scores higher on coding benchmarks: 46.5 versus 30.9 in the Noometry coding category.
How many benchmarks do DeepSeek-V3.2-Exp and Llama 2-13B share?
20 benchmarks have published results for both models. DeepSeek-V3.2-Exp has 49 scored results on Noometry and Llama 2-13B has 32.