Free Video AI Officially Launches, Dual Versions for Different Creation Needs
Agnes Video, developed by Agnes Studio (formerly known as “凹非寺”), has been upgraded to version 2.5 and is now live on the Pavo creation platform. The new series comprises two variants: Agnes Video 2.5 Flash and Agnes Video 2.5, both freely accessible to creators.
Key facts at a glance:
- Release time: Late August 2026
- New version: Agnes Video 2.5 series (Flash and Pro sub-versions)
- Pricing strategy: Flash version completely free; Pro version offers 200 daily credits
- Availability: Live on Pavo platform
- Credit rules: Pro version consumes 2 credits per second, max 100 seconds per generation
Infinite Canvas and Short Film Mode Reduce Creative Barriers
Pavo offers two standout features for the new model series. The Infinite Canvas功能 allows creators to add nodes of text, images, videos, audio, and more, with links between nodes and duplicate branches for variant testing. In the “牛来” short film production, for instance, creators combine 3D character assets with background images to generate initial frames, then continue generating video from satisfactory first frames.
For multi-character scenes, the platform’s 3D Director’s Desk enables fine-grained control of character positioning, poses, and proportional scaling—preventing model default behaviors that could cause inconsistency.
A second flagship feature is the Short Film Mode, which guides users through five stages automatically:
- Requirement confirmation
- Script outline
- Character, scene, and prop design
- Storyboard frames and key frames
- Scene video合成 and final cut
Most generation tasks run autonomously in AI; user effort is primarily confined to confirming outputs at each stage.
Performance Parameters: 2.5 Version Shows Clear Improvement
| Metric | Agnes Video 2.0 | Agnes Video 2.5 Flash | Agnes Video 2.5 Pro |
|---|---|---|---|
| Pricing | Paid | Free | 200 credits/day (2 credits/sec) |
| Video length | Not specified | Unlimited | Max 100 seconds |
| Resolution | Not specified | Not specified | Up to 2K |
| Reference assets | Basic support | Unlimited | Multi-image + audio/video reference |
| AA ranking | Lower | Not separately ranked | Global mainstream tier |
| Cost per second | — | Free | ~RMB 0.15 |
The surprising metric: the Pro version, previously expensive, now costs only RMB 0.15 per second. The free Flash version’s output quality has advanced sufficiently to reach Artificial Analysis榜单的主流模型水平—大幅前进 versus the 2.0 version.
Who Should Use It Now? Who Should Wait?
Immediate adoption recommended for:
- Short-form video creators needing high-frequency testing and variant iteration
- Small teams/personal creators with limited budgets yet requiring 2K commercial content
- Beginners unfamiliar with prompt engineering, who benefit from guided short film workflows
Better to wait if you require:
- Professional film teams needing videos longer than 100 seconds (Pro version cap)
- Advertisers demanding extremely precise style reproduction (specific metrics unconfirmed)
Final Thoughts
When “free” becomes a performance guarantee rather than a compromise, video AI competition shifts from “single-shot aesthetic appeal” to “full-process controllability”. The free Flash version’s launch marks not just a cost breakthrough, but a signal of video AI maturity—where creativity, finally, becomes the sole bottleneck.