All benchmarksIndependent evaluation · Epoch AI
SuperGLUE
Each model’s result as Epoch AI’s Benchmarking Hub publishes it, run by Epoch or collected from the source it cites. Facthory has not run them.
Data: Epoch AI, ‘Capabilities & benchmarking’, epoch.ai/benchmarks, licensed CC BY 4.0. Read on 16 Nov 2026.
| # | Model | Organization | Result | Evidence |
|---|---|---|---|---|
| 1 | T5-11B Open weights (unrestricted) | 88.9% | Log or source | |
| 2 | T5-3B Open weights (unrestricted) | 86.4% | Log or source | |
| 3 | Switch-Large | – | 84.7% | – |
| 4 | T5-Large | – | 82.7% | – |
| 5 | T5-Base | – | 75.1% | – |
| 6 | Switch-Base | – | 73.3% | – |
| 7 | text-davinci-001 API access | OpenAI | 71.8% | – |
| 8 | T5-Small | – | 63.3% | Log or source |
