ServImage - Deliverable Acceptance: leaderboard
ServImage (2026) evaluates image generation and editing models on real commercial imaging-service tasks in portrait, product and digital design, judged by whether a human client would pay for the delivered images. This board is the share of individual delivered images approved by human payment decisions, in percent; higher is better.
Metric: Deliverable acceptance rate (%): share of individual generated or edited images approved by human payment decisions, standard settlement scenario; higher is better. Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 16 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Nano Banana Pro | 79.88 | |
| 2 | GPT Image 1 | 46.38 | |
| 3 | Seedream-3.0 | 41.54 | |
| 4 | Seedream 4.0 | 40.18 | |
| 5 | gpt-image-1-mini | 37.62 | |
| 6 | Nano Banana | 34.56 | |
| 7 | Ideogram-v3 | 31.67 | |
| 8 | Qwen-Image | 30.4 | |
| 9 | Kling-v2 | 28.85 | |
| 10 | Imagen 4.0 | 28.1 | |
| 11 | MJ-Relax-Imagine (ServImage checkpoint unspecified) | 28.01 | |
| 12 | MJ-Relax-Edits (ServImage checkpoint unspecified) | 25.9 | |
| 13 | FLUX-Kontext-Pro | 14.18 | |
| 14 | FLUX 1.1 Pro | 8.85 | |
| 15 | SeedEdit 3.0 | 6.77 |
Interactive version: theaggregate.ai/benchmark?slug=servimage-deliverable-acceptance · How It Works · Data refreshed daily, snapshot 2026-11-22.