Vector Eval - IFEval: leaderboard
Vector Institute Inspect Evals run of IFEval, testing strict adherence to verifiable natural-language instructions and formatting constraints.
Metric: Final Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 1.5 Pro | 89.82 | #445 |
| 2 | Claude 3.5 Sonnet | 89.58 | #350 |
| 3 | GPT-4o | 87.8 | #343 |
| 4 | Llama 3.1 70B Instruct | 86.05 | #561 |
| 5 | GPT-4o Mini | 84.19 | #599 |
| 6 | Mistral Large 2 (Jul) | 82.85 | #484 |
| 7 | Command-R+ | 77.8 | #891 |
| 8 | Gemini 1.5 Flash | 76.81 | #618 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=vector-eval-ifeval · How It Works · Data refreshed daily, snapshot 2026-10-20.