SaaS-Bench - Independent Media Creation: leaderboard
Metric: Checkpoint score (%) on the independent media creation domain (multimodal), weighted-checkpoint score over the final application states, browser-use agent loop over live self-hosted SaaS apps (DOM and screenshot observations, restricted browser actions, no JavaScript or backend access), same step budget, timeout and failure cap for every model; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 | 52.8 |
| 2 | Claude Opus 4.7 | 46.3 |
| 3 | GPT-5.5 (High) | 45.2 |
| 4 | Gemini 3.1 Pro (Preview) | 44.7 |
| 5 | Seed 2.0 Pro | 42.9 |
| 6 | GPT-5.4 (High) | 41.7 |
| 7 | Qwen 3.6 Plus | 41.3 |
| 8 | Kimi K2.6 | 39.5 |
| 9 | Gemini 3.5 Flash (High) | 36.9 |
| 10 | Claude Sonnet 4.6 | 33.8 |
| 11 | Kimi K2.5 | 28.5 |
Interactive version: theaggregate.ai/benchmark?slug=saas-bench-independent-media-creation · How It Works · Data refreshed daily, snapshot 2026-10-07.