May 2026
Can AI agents do real-world video post-production work?
We gave the 9 best frontier models 100 expert-authored tasks across the four stages of video post-production. The best agent tops out at 32%. Human experts scored 89%.
100
Tasks
20
Industry experts
7
Frontier models
4
Task families
Why this benchmark exists
Verification is not here for free.
RLVR - reinforcement learning with verifiable rewards - works in math and code because centuries of humanistic work built the verifiers; the bill was paid before we got there. Creative work hasn't paid that bill. AgenticVBench is what paying it looks like in film.
It also measures the sim2real gap: the distance between how agents score on tidy lab benchmarks and how they hold up on real post-production work. Here that gap is stark - the best frontier agent scores 32%, human experts 89%.
Read the full essay →Leaderboard preview
Top 5 model × harness combinations.
| Rank | Agent | Avg | Repurpose | Seq | Repair | Assembly |
|---|---|---|---|---|---|---|
| · | Human expertsreference | 88.5% | 95% | 90% | 88% | 81% |
| 1 | Claude Fable 5· Claude Code | 32.4%± 5.1 | 29% | 23% | 31% | 46% |
| 2 | GPT-5.5· Codex | 31.0%± 4.0 | 30% | 26% | 30% | 38% |
| 3 | GPT-5.5· OpenCode | 27.4%± 3.5 | 27% | 20% | 27% | 37% |
| 4 | Gemini 3.1 Pro· OpenCode | 23.8%± 3.7 | 23% | 19% | 20% | 33% |
| 5 | MiniMax-M3· OpenCode | 22.7%± 3.2 | 23% | 12% | 18% | 37% |
What the bench tests
Four task families spanning the real-world post-production workflow.
Authored by 20 industry experts averaging 6 years of post-production experience. Tasks span 30 minutes to one week of human work.
Assembly
18 tasks
35
pp gap
Given a storyboard with 3–6 slots and a shuffled pool of candidate clips, select the clip that matches each slot.
Repair
18 tasks
57
pp gap
Given a video with defects (frozen scene, scene swap, color drift, or audio noise), localize them and produce a fixed cut.
Sequencing
28 tasks
61
pp gap
Given a brief story overview and a shuffled set of clips, recover the correct narrative order.
Repurpose
36 tasks
65
pp gap
Given 4-150 minutes of source video and a creative brief, repurpose it into a short deliverable that follows the brief and preserves the story.
The harness finding
The harness matters as much as the model.
Holding the model fixed and varying the harness shifts GPT-5.5's Assembly score by 20 percentage points, comparable to the gap between adjacent models on the leaderboard.
Most benchmarks today are still model-based. The data here says that's wrong. Agent performance is determined by both the model and the scaffolding around it. Reporting only the model misses the larger story.
Agent = model × harness.
GPT-5.5 on Assembly · score by harness
Same model. 20-point swing.
Cite this work
Citation
If you find AgenticVBench useful in your research, please consider citing the paper.
@article{cao2026agenticvbench,
title={AgenticVBench: Can AI Agents Complete Real-World Post-Production Tasks?},
author={Cao, Zongheng and Zheng, Yi and Song, Rui and Hu, Xinyu},
journal={arXiv preprint arXiv:2605.27705},
year={2026}
}