HELM Classic - NarrativeQA: leaderboard
The "NarrativeQA" subtask of HELM Classic. A scenario from Stanford HELM Classic, the holistic evaluation suite scoring language models across accuracy, robustness and calibration on a broad set of tasks.
Metric: F1 (%). Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 66 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 2 70B | 76.99 | #1201 |
| 2 | LLaMA-65B | 75.49 | #1303 |
| 3 | LLaMA-30B | 75.25 | #1423 |
| 4 | Llama 2 13B | 74.4 | #1436 |
| 5 | Palmyra X (43B) | 74.19 | #1317 |
| 6 | mpt-30B Instruct | 73.29 | #1470 |
| 7 | mpt-30B | 73.15 | #1476 |
| 8 | text-davinci-002 | 72.72 | #1225 |
| 9 | text-davinci-003 | 72.71 | #1146 |
| 10 | Mistral-7B-v0.1 | 71.65 | #1326 |
| 11 | LLaMA-13B | 71.12 | #1516 |
| 12 | GLM-130B | 70.59 | #1568 |
| 13 | Llama 2 7B | 69.12 | #1520 |
| 14 | vicuna-13B-v1.3 | 69.07 | #1447 |
| 15 | davinci | 68.69 | #1553 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-classic-narrativeqa · How It Works · Data refreshed daily, snapshot 2026-10-20.