GPT-OSS-120B (Reasoning): benchmark results
Provider: OpenAI. Released 2025-08-05. Access: Open.
Aggregate Rating 1565 ± 14, rank #885 of 2287 rated models, from 146 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SEA-HELM (English) - LiveCodeBench v6 | 68.18 | Normalized Score | 91.7 |
| BALSAM - Question Answering | 73.39 | Overall score (0-100, LLM-judged generation and multiple cho | 88.7 |
| SEA-HELM (English) - MathArena | 45.45 | Normalized Score | 88.3 |
| BALSAM - Overall | 55.91 | Mean of category overall scores (0-100) | 82.8 |
| SEA-HELM (Filipino) - SEA-Safeguard | 75.89 | Normalized Score | 81.1 |
| BALSAM - Entailment | 79.5 | Overall score (0-100, LLM-judged generation and multiple cho | 80.6 |
| BALSAM - Text Classification | 39.35 | Overall score (0-100, LLM-judged generation and multiple cho | 80 |
| SEA-HELM (Tamil) - SEA-MT-Bench (LLM Judge) | 79.82 | Normalized Score | 77 |
| BALSAM - Program Execution | 93.78 | Overall score (0-100, LLM-judged generation and multiple cho | 75.8 |
| BALSAM - Sequence Tagging | 40.38 | Overall score (0-100, LLM-judged generation and multiple cho | 75 |
| SEA-HELM (Thai) - SEA-MT-Bench (LLM Judge) | 83.26 | Normalized Score | 74.3 |
| BALSAM - Reading Comprehension | 56.77 | Overall score (0-100, LLM-judged generation and multiple cho | 74.2 |
Interactive version: theaggregate.ai/model?slug=gpt-oss-120b-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-22.