Llama 3.3 70B Instruct
Llama 3.3 70B Instruct is behind the leaders: #288 of 406 by SIscope score. Strongest at chat quality, weaker at math. Cheaper than 91% of priced models. Weights are open, so you can run it yourself.
Shape
Each spoke is a category, 0 at the centre, 100 at the edge.
Category scores
Percentiles: 90 means better than about 90% of tested models.
Text arena over time
Rating (solid) and rank (dashed) at each Arena update we have: 3 of 16 snapshots.
Independent test runs
Run by Epoch AI on the same harness for every model. Bars show the share of questions answered correctly.
Human-vote arenas
People compare two anonymous answers and pick the better one. Scores work like chess ratings.
| Arena | Rank | Score | Range (95%) | Votes | Updated |
|---|---|---|---|---|---|
| Text arena ↗ Everyday chat, blind human votes | #279 of 419 | 1274 | 1271–1278 | 54,493 | Oct 28, 2026 |
| WebDev arena ↗ Building web apps, blind human votes | Not on this board yet | ||||
| Vision arena ↗ Questions about images, blind human votes | Not on this board yet | ||||
| Agent arena ↗ Real multi-step agent sessions | Not on this board yet | ||||
Source: Arena (formerly LMArena) leaderboard dataset, CC BY 4.0. Fetched Nov 4, 2026.
Price
List price from the OpenRouter models API, Nov 4, 2026: meta-llama/llama-3.3-70b-instruct on OpenRouter ↗. Providers can charge differently.