Up to four models
GPT-5 vs gpt-oss-20b
On the index (GPT-4 = 100), GPT-5 scores 149 and gpt-oss-20b 128. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
GPT-5
149133–164
WeightsClosed
ReleasedAug 2025
Price in / out$1.25 / $10
Context400K
Core tests taken6 of 9
gpt-oss-20b
128112–144
WeightsOpen
ReleasedAug 2025
Price in / out$0.018 / $0.090
Context131K
Core tests taken1 of 9
GPT-5 leads gpt-oss-20b by 21 points, ahead on 1 of 1 shared core tests, at 95× the price.
Test by testthird-party results
GPT-5gpt-oss-20b
Every result as a table
| Test | GPT-5 | gpt-oss-20b |
|---|---|---|
| Reasoning & knowledge | ||
| GPQA Diamond (Epoch AI) | 86.2% | 60.8% |
| GPQA Diamond (Vals AI) | 85.6% | 68.9% |
| MMLU Pro | 86.5% | 71.6% |
| Maths | ||
| AIME-style maths (OTIS mock) | 91.4% | 65.3% |
| MATH 500 | 96.0% | 94.2% |
| MGSM | 92.8% | 89.0% |
| Coding | ||
| Terminal-Bench 2.0 (Terminus 2 runs) | 35.2% | 3.1% |
| LiveCodeBench | 85.9% | 80.4% |
| WeirdML | 60.7% | 40.9% |
| Agents | ||
| DTBench | 90.7% | 68.0% |
| LMCA | 40.0% | 14.5% |
| Professional work | ||
| CorpFin v2 | 61.1% | 53.1% |
| LegalBench | 86.0% | 70.9% |
| TaxEval v2 | 73.4% | 63.7% |
Task length · METR42 sec
Tests at least two of these models have taken. Bold rows are the core tests.