Model comparison
GPT-5.2 vs gpt-oss-120b
GPT-5.2 is the stronger model overall, scoring 54.1 to 36.3 on the Noometry Index. gpt-oss-120b costs 69× less per token, which makes it the better buy when GPT-5.2's lead doesn't matter for your workload.
Last verified . 35 shared benchmarks.
Summary
- They share 35 benchmarks with published results for both. GPT-5.2 scores higher in 9 categories and gpt-oss-120b in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.2 leads 50.2 to 20.0.
- The biggest single-benchmark swing is SWE-bench Verified (bash only): 72.8% for GPT-5.2 and 26% for gpt-oss-120b.
- gpt-oss-120b is cheaper at $0.037 / $0.17 per million input/output tokens, against $1.75 / $14 for GPT-5.2.
- GPT-5.2 accepts more context: 400K tokens versus 131K.
- gpt-oss-120b has downloadable open weights; the other is API-only.
Side by side
| GPT-5.2 | gpt-oss-120b | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 54.1 | 36.3 |
| Released | 2025-12-11 | 2025-08-05 |
| Weights | Proprietary | Open |
| Context window | 400K | 131K |
| Max output | 128K | 41K |
| Input $ / M tokens | $1.75 | $0.037 |
| Output $ / M tokens | $14 | $0.17 |
| Results tracked | 67 | 48 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.2 leads
GPT-5.2: 51.6 (#37), gpt-oss-120b: 33.5 (#256)
| Benchmark | GPT-5.2 | gpt-oss-120b |
|---|---|---|
| SWE-bench Verified (bash only) | 72.8% | 26% |
| WeirdML | 72.2% | 48.2% |
| LMArena Coding | 1447 | 1380 |
| ALE-Bench | 1,294 | 575.62 |
| AlgoTune | 2.05 | 1.41 |
| SWE-bench Verified | 73.8% | — |
| Aider Polyglot | — | 41.8% |
| LMArena WebDev | 1416 | — |
| SWE-bench Multilingual | 66.7% | — |
| SciCode | — | 36% |
| GSO | 27.4% | — |
Agentic & Tool Use GPT-5.2 leads
GPT-5.2: 40.2 (#24), gpt-oss-120b: 12.2 (#153)
| Benchmark | GPT-5.2 | gpt-oss-120b |
|---|---|---|
| Terminal-Bench | 64.9% | 18.7% |
| METR Time Horizons | 75.3% | 56.6% |
| Vending-Bench 2 | 3,591 | -21.53 |
| APEX-Agents | — | 4.4% |
| Berkeley Function Calling Leaderboard | 55.9% | — |
| GDPval | 49.7% | — |
| Remote Labor Index | 2.5% | — |
| τ²-bench Airline | 83% | — |
| τ²-bench Banking | 32.2% | — |
| τ²-bench Retail | 81.6% | — |
| τ²-bench Telecom | 89.7% | — |
| DeepResearch Bench | 41.1% | — |
| LMArena Search | 1207 | — |
Reasoning GPT-5.2 leads
GPT-5.2: 50.2 (#35), gpt-oss-120b: 20.0 (#245)
| Benchmark | GPT-5.2 | gpt-oss-120b |
|---|---|---|
| SimpleBench | 45.8% | 22.1% |
| Kagi LLM Benchmark | 73.3% | 58.6% |
| Chess Puzzles | 49% | 20% |
| LMArena Hard Prompts | 1445 | 1364 |
| Mystery Game Puzzles | 23% | 2% |
| DTBench | 90.9% | 76.3% |
| LMCA | 43.9% | 22.1% |
| Epoch Capabilities Index | 153.45 | 139.93 |
| ARC-AGI-2 | 52.9% | — |
| NYT Connections (extended) | 83.6% | — |
| ARC-AGI-1 | 86.2% | — |
| CritPt | — | 1.1% |
| EnigmaEval | 10.4% | — |
| EBR-Bench | 23% | — |
| Surface Evolver Bench | — | 25% |
| ForecastBench | 60.1 | — |
Math GPT-5.2 leads
GPT-5.2: 60.0 (#38), gpt-oss-120b: 52.5 (#50)
| Benchmark | GPT-5.2 | gpt-oss-120b |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 96.1% | 88.9% |
| LMArena Math | 1440 | 1389 |
| FrontierMath (Tiers 1-3) | 67.4% | — |
| FrontierMath Tier 4 | 31.7% | — |
| MathArena Final-Answer Competitions | 72% | — |
| ProofBench | 15% | — |
| Omni-MATH | — | 68.8% |
| FrontierMath (Feb 2025 set) | 40.7% | — |
| FrontierMath Tier 4 (v1) | 18.8% | — |
Knowledge GPT-5.2 leads
GPT-5.2: 59.3 (#32), gpt-oss-120b: 42.4 (#96)
| Benchmark | GPT-5.2 | gpt-oss-120b |
|---|---|---|
| GPQA Diamond | 91.4% | 75.8% |
| Vectara Hallucination Rate | 8.4% | 14.2% |
| LMArena Expert | 1445 | 1356 |
| Humanity's Last Exam | 27.8% | — |
| SimpleQA Verified | 37.1% | — |
| MMLU-Pro | — | 79.5% |
| Confabulations | — | 15.7% |
| GPQA (HELM) | — | 68.4% |
Multimodal Not comparable
GPT-5.2: 51.3 (#7), gpt-oss-120b: —
| Benchmark | GPT-5.2 | gpt-oss-120b |
|---|---|---|
| LMArena Vision | 1268 | — |
| VPCT | 84% | — |
| Furniture Assembly | 38.3% | — |
| LMArena Document | 1405 | — |
Multilingual GPT-5.2 leads
GPT-5.2: 53.4 (#67), gpt-oss-120b: 48.0 (#147)
| Benchmark | GPT-5.2 | gpt-oss-120b |
|---|---|---|
| LMArena Non-English | 1425 | 1351 |
| LMArena Chinese | 1460 | 1385 |
| LMArena French | 1455 | 1369 |
| LMArena German | 1448 | 1353 |
| LMArena Japanese | 1420 | 1331 |
| LMArena Korean | 1392 | 1282 |
| LMArena Russian | 1440 | 1343 |
| LMArena Spanish | 1433 | 1389 |
Instruction Following GPT-5.2 leads
GPT-5.2: 74.7 (#89), gpt-oss-120b: 69.3 (#173)
| Benchmark | GPT-5.2 | gpt-oss-120b |
|---|---|---|
| LMArena Instruction Following | 1417 | 1318 |
| IFEval | — | 83.6% |
Long Context GPT-5.2 leads
GPT-5.2: 44.0 (#78), gpt-oss-120b: 31.4 (#278)
| Benchmark | GPT-5.2 | gpt-oss-120b |
|---|---|---|
| LMArena Longer Query | 1428 | 1319 |
| Fiction.LiveBench | — | 44.4% |
| CL-bench | 18.2% | — |
Writing & Preference GPT-5.2 leads
GPT-5.2: 66.8 (#32), gpt-oss-120b: 46.5 (#217)
| Benchmark | GPT-5.2 | gpt-oss-120b |
|---|---|---|
| LMArena Text | 1439 | 1365 |
| LMArena Creative Writing | 1401 | 1275 |
| EQ-Bench Creative Writing | 1703 | 961 |
| LMArena Multi-Turn | 1458 | 1340 |
| Short-Story Creative Writing | — | 77.1% |
| WildBench | — | 84.5% |
Frequently asked questions
Is GPT-5.2 better than gpt-oss-120b?
GPT-5.2 is the stronger model overall, scoring 54.1 to 36.3 on the Noometry Index. gpt-oss-120b costs 69× less per token, which makes it the better buy when GPT-5.2's lead doesn't matter for your workload.
Which is cheaper, GPT-5.2 or gpt-oss-120b?
gpt-oss-120b is cheaper. It lists at $0.037 per million input tokens and $0.17 per million output tokens; GPT-5.2 lists at $1.75 and $14.
Is GPT-5.2 or gpt-oss-120b better for coding?
GPT-5.2 scores higher on coding benchmarks: 51.6 versus 33.5 in the Noometry coding category.
Which has the bigger context window?
GPT-5.2 does, with 400K tokens against 131K.
How many benchmarks do GPT-5.2 and gpt-oss-120b share?
35 benchmarks have published results for both models. GPT-5.2 has 67 scored results on Noometry and gpt-oss-120b has 48.