Nova Pro (v1): benchmark results
Amazon Nova Pro v1, the balanced flagship multimodal model of the first Nova generation (December 2024). Provider: Amazon. Released 2024-12-03. Access: API.
Aggregate Rating 1583 ± 11, rank #497 of 1656 rated models, from 292 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| BALSAM - Reading Comprehension | 61.09 | Overall score (0-100, LLM-judged generation and multiple cho | 96.3 |
| Evals for Every Language - Greek (el) | 78.2 | Average Score (%) | 95.5 |
| Vectara Hallucination Leaderboard | 94.9 | Factual Consistency Rate (%) | 90.7 |
| Evals for Every Language - Armenian (hy) | 74.14 | Average Score (%) | 90.6 |
| BALSAM - Entailment | 86.25 | Overall score (0-100, LLM-judged generation and multiple cho | 88.9 |
| AGC-Bench - futuregen | 1.18 | Dataset z-score | 88.8 |
| TextArena Balanced Subset | 28.46 | Trueskill (rating) | 85 |
| AGC-Bench - grapheval_ai_researcher | 0.67 | Dataset z-score | 81.5 |
| BALSAM - Question Answering | 72.87 | Overall score (0-100, LLM-judged generation and multiple cho | 81.5 |
| Evals for Every Language - Awadhi (awa) | 71.02 | Average Score (%) | 81.5 |
| Evals for Every Language - Ukrainian (uk) | 76.52 | Average Score (%) | 81.5 |
| Evals for Every Language - Classification | 95.21 | Average Score (%) | 80.9 |
Interactive version: theaggregate.ai/model?slug=nova-pro-v1 · How It Works · Data refreshed daily, snapshot 2026-10-20.