Model comparison
gpt-oss-20b vs Llama 3.2 1B
gpt-oss-20b is the stronger model overall, scoring 32.1 to 19.9 on the Noometry Index.
Last verified . 18 shared benchmarks.
MetaSummary
- They share 18 benchmarks with published results for both. gpt-oss-20b scores higher in 8 categories and Llama 3.2 1B in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where gpt-oss-20b leads 37.6 to 9.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 65.3% for gpt-oss-20b and 0.6% for Llama 3.2 1B.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $0.027 / $0.20 for Llama 3.2 1B.
- gpt-oss-20b accepts more context: 131K tokens versus 60K.
Side by side
| gpt-oss-20b | Llama 3.2 1B | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 32.1 | 19.9 |
| Released | 2025-08-05 | 2024-09-24 |
| Weights | Open | Open |
| Context window | 131K | 60K |
| Max output | 16K | 54K |
| Input $ / M tokens | $0.018 | $0.027 |
| Output $ / M tokens | $0.09 | $0.20 |
| Results tracked | 34 | 22 |
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 3.2 1B: 20.6 (#333)
| Benchmark | gpt-oss-20b | Llama 3.2 1B |
|---|---|---|
| LMArena Coding | 1307 | 1071 |
| SciCode | 34.4% | — |
| WeirdML | 40.9% | — |
| BigCodeBench Instruct | — | 8.2% |
| BigCodeBench Complete | — | 11.3% |
| ALE-Bench | 566.05 | — |
Agentic & Tool Use Llama 3.2 1B leads
gpt-oss-20b: 9.2 (#146), Llama 3.2 1B: 14.5 (#142)
| Benchmark | gpt-oss-20b | Llama 3.2 1B |
|---|---|---|
| Terminal-Bench | 3.4% | — |
| Berkeley Function Calling Leaderboard | — | 10.8% |
| BALROG | — | 6.6% |
Reasoning gpt-oss-20b leads
gpt-oss-20b: 18.8 (#251), Llama 3.2 1B: 15.6 (#304)
| Benchmark | gpt-oss-20b | Llama 3.2 1B |
|---|---|---|
| Chess Puzzles | 4% | 0% |
| LMArena Hard Prompts | 1274 | 1044 |
| Epoch Capabilities Index | 137.81 | 101.95 |
| Kagi LLM Benchmark | 53.2% | — |
| CritPt | 1.4% | — |
| DTBench | 68% | — |
| LMCA | 14.5% | — |
Math gpt-oss-20b leads
gpt-oss-20b: 37.6 (#93), Llama 3.2 1B: 9.7 (#308)
| Benchmark | gpt-oss-20b | Llama 3.2 1B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 65.3% | 0.6% |
| LMArena Math | 1317 | 1086 |
| Omni-MATH | 56.5% | — |
Knowledge gpt-oss-20b leads
gpt-oss-20b: 34.6 (#186), Llama 3.2 1B: 7.2 (#307)
| Benchmark | gpt-oss-20b | Llama 3.2 1B |
|---|---|---|
| GPQA Diamond | 60.8% | 23.9% |
| LMArena Expert | 1258 | 1007 |
| MMLU-Pro | 74% | — |
| GPQA (HELM) | 59.4% | — |
Multilingual gpt-oss-20b leads
gpt-oss-20b: 42.4 (#192), Llama 3.2 1B: 23.9 (#287)
| Benchmark | gpt-oss-20b | Llama 3.2 1B |
|---|---|---|
| LMArena Non-English | 1269 | 973 |
| LMArena Chinese | 1311 | 958 |
| LMArena German | 1249 | 1014 |
| LMArena Russian | 1279 | 941 |
| LMArena Japanese | 1245 | — |
| LMArena Korean | 1239 | — |
| LMArena Spanish | 1267 | — |
Instruction Following gpt-oss-20b leads
gpt-oss-20b: 62.0 (#235), Llama 3.2 1B: 52.6 (#285)
| Benchmark | gpt-oss-20b | Llama 3.2 1B |
|---|---|---|
| LMArena Instruction Following | 1236 | 1031 |
| IFEval | 73.2% | — |
Long Context gpt-oss-20b leads
gpt-oss-20b: 37.9 (#203), Llama 3.2 1B: 31.9 (#269)
| Benchmark | gpt-oss-20b | Llama 3.2 1B |
|---|---|---|
| LMArena Longer Query | 1251 | 1050 |
Writing & Preference gpt-oss-20b leads
gpt-oss-20b: 35.8 (#260), Llama 3.2 1B: 21.4 (#305)
| Benchmark | gpt-oss-20b | Llama 3.2 1B |
|---|---|---|
| LMArena Text | 1288 | 1055 |
| LMArena Creative Writing | 1201 | 1033 |
| EQ-Bench Creative Writing | 666 | 200 |
| LMArena Multi-Turn | 1268 | 1030 |
| WildBench | 73.7% | — |
Frequently asked questions
Is gpt-oss-20b better than Llama 3.2 1B?
gpt-oss-20b is the stronger model overall, scoring 32.1 to 19.9 on the Noometry Index.
Which is cheaper, gpt-oss-20b or Llama 3.2 1B?
gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; Llama 3.2 1B lists at $0.027 and $0.20.
Is gpt-oss-20b or Llama 3.2 1B better for coding?
gpt-oss-20b scores higher on coding benchmarks: 37.1 versus 20.6 in the Noometry coding category.
Which has the bigger context window?
gpt-oss-20b does, with 131K tokens against 60K.
How many benchmarks do gpt-oss-20b and Llama 3.2 1B share?
18 benchmarks have published results for both models. gpt-oss-20b has 34 scored results on Noometry and Llama 3.2 1B has 22.