Model comparison
gpt-oss-20b vs Llama 2-70B
gpt-oss-20b is the stronger model overall, scoring 32.1 to 24.1 on the Noometry Index.
Last verified . 20 shared benchmarks.
MetaSummary
- They share 20 benchmarks with published results for both. gpt-oss-20b scores higher in 8 categories and Llama 2-70B in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where gpt-oss-20b leads 37.6 to 7.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 65.3% for gpt-oss-20b and 0% for Llama 2-70B.
Side by side
| gpt-oss-20b | Llama 2-70B | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 32.1 | 24.1 |
| Released | 2025-08-05 | 2023-07-18 |
| Weights | Open | Open |
| Context window | 131K | — |
| Max output | 16K | — |
| Input $ / M tokens | $0.018 | — |
| Output $ / M tokens | $0.09 | — |
| Results tracked | 34 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding gpt-oss-20b leads
gpt-oss-20b: 37.1 (#185), Llama 2-70B: 30.8 (#281)
| Benchmark | gpt-oss-20b | Llama 2-70B |
|---|---|---|
| LMArena Coding | 1307 | 1079 |
| SciCode | 34.4% | — |
| WeirdML | 40.9% | — |
| ALE-Bench | 566.05 | — |
Agentic & Tool Use Not comparable
gpt-oss-20b: 9.2 (#146), Llama 2-70B: —
| Benchmark | gpt-oss-20b | Llama 2-70B |
|---|---|---|
| Terminal-Bench | 3.4% | — |
Reasoning gpt-oss-20b leads
gpt-oss-20b: 18.8 (#251), Llama 2-70B: 13.7 (#324)
| Benchmark | gpt-oss-20b | Llama 2-70B |
|---|---|---|
| LMArena Hard Prompts | 1274 | 1073 |
| DTBench | 68% | 41.6% |
| Epoch Capabilities Index | 137.81 | 113.59 |
| Kagi LLM Benchmark | 53.2% | — |
| CritPt | 1.4% | — |
| Chess Puzzles | 4% | — |
| LMCA | 14.5% | — |
| BIG-Bench Hard | — | 64.9% |
| CommonsenseQA 2.0 | — | 50% |
| ForecastBench | — | 51.4 |
| HellaSwag | — | 85.3% |
| LAMBADA | — | 78.9% |
| PIQA | — | 82.8% |
| WinoGrande | — | 80.2% |
Math gpt-oss-20b leads
gpt-oss-20b: 37.6 (#93), Llama 2-70B: 7.5 (#320)
| Benchmark | gpt-oss-20b | Llama 2-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 65.3% | 0% |
| LMArena Math | 1317 | 1091 |
| Omni-MATH | 56.5% | — |
| MATH Level 5 | — | 3.3% |
| GSM8K | — | 69.6% |
Knowledge gpt-oss-20b leads
gpt-oss-20b: 34.6 (#186), Llama 2-70B: 7.4 (#305)
| Benchmark | gpt-oss-20b | Llama 2-70B |
|---|---|---|
| GPQA Diamond | 60.8% | 26.3% |
| LMArena Expert | 1258 | 1039 |
| MMLU-Pro | 74% | — |
| GPQA (HELM) | 59.4% | — |
| ARC (AI2) Challenge | — | 78.3% |
| BoolQ | — | 88.6% |
| MMLU | — | 69.9% |
| OpenBookQA | — | 60.2% |
| TriviaQA | — | 87.6% |
Multilingual gpt-oss-20b leads
gpt-oss-20b: 42.4 (#192), Llama 2-70B: 27.9 (#269)
| Benchmark | gpt-oss-20b | Llama 2-70B |
|---|---|---|
| LMArena Non-English | 1269 | 1045 |
| LMArena Chinese | 1311 | 994 |
| LMArena German | 1249 | 1041 |
| LMArena Japanese | 1245 | 927 |
| LMArena Korean | 1239 | 964 |
| LMArena Russian | 1279 | 1083 |
| LMArena Spanish | 1267 | 1142 |
| LMArena French | — | 1089 |
Instruction Following gpt-oss-20b leads
gpt-oss-20b: 62.0 (#235), Llama 2-70B: 55.1 (#273)
| Benchmark | gpt-oss-20b | Llama 2-70B |
|---|---|---|
| LMArena Instruction Following | 1236 | 1071 |
| IFEval | 73.2% | — |
Long Context gpt-oss-20b leads
gpt-oss-20b: 37.9 (#203), Llama 2-70B: 32.2 (#265)
| Benchmark | gpt-oss-20b | Llama 2-70B |
|---|---|---|
| LMArena Longer Query | 1251 | 1062 |
Writing & Preference gpt-oss-20b leads
gpt-oss-20b: 35.8 (#260), Llama 2-70B: 32.5 (#274)
| Benchmark | gpt-oss-20b | Llama 2-70B |
|---|---|---|
| LMArena Text | 1288 | 1115 |
| LMArena Creative Writing | 1201 | 1075 |
| LMArena Multi-Turn | 1268 | 1088 |
| EQ-Bench Creative Writing | 666 | — |
| WildBench | 73.7% | — |
Frequently asked questions
Is gpt-oss-20b better than Llama 2-70B?
gpt-oss-20b is the stronger model overall, scoring 32.1 to 24.1 on the Noometry Index.
Is gpt-oss-20b or Llama 2-70B better for coding?
gpt-oss-20b scores higher on coding benchmarks: 37.1 versus 30.8 in the Noometry coding category.
How many benchmarks do gpt-oss-20b and Llama 2-70B share?
20 benchmarks have published results for both models. gpt-oss-20b has 34 scored results on Noometry and Llama 2-70B has 35.