TECCI - Instruction Following: leaderboard
Metric: Instruction-following human success rate (%; five trained raters score each edit 1-5 and a criterion succeeds when the mean is at least 4.5; 1,315 sampled pairs: 265 of the Instruction Rich Challenge Set and 1,050 of the Gemini Generated Instruction Set). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 5 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Grok Imagine Pro | 46.9 |
| 2 | Nano Banana Pro | 46 |
| 3 | Nano Banana 2 (Gemini 3.1 Flash Image Preview) | 45.6 |
| 4 | GPT Image 1.5 | 42 |
| 5 | Seedream 5.0 Lite | 41.4 |
Interactive version: theaggregate.ai/benchmark?slug=tecci-instruction-following · How It Works · Data refreshed daily, snapshot 2026-09-26.