Timeline Studio Guide: Local-First AI Video Editing in the Browser
Answer in brief
Timeline Studio is a local-first browser video editor with a multi-track timeline, WebGPU-assisted AI tools, multilingual voiceovers, automatic captions, and deterministic offline export. Some optional generation connectors use browser-direct or loopback services, so users should verify the active workflow before handling sensitive media and must clear all model, dataset, and media rights before publication or commercial use. The official source does not publish a selectable model ID for this product.
Key facts at a glance
| Product / model | Current ID or version | Use case | Evidence |
|---|---|---|---|
| timeline-studio | Official source does not specify a selectable model ID | Confirm the current product surface | Official source Official source Official source |
Failure modes and verification
| Failure mode | Verification action |
|---|---|
| Stale model or version reference | Compare the model name and ID with the official source before release. |
| Unstructured or incomplete output | Validate the response against the documented contract and a deterministic fixture. |
| Unverified factual claim | Keep the claim qualified or remove the claim when the official source does not support it. |
FAQ
Does Timeline Studio run entirely offline?
Not necessarily. The editor is officially described as local-first, and several AI workflows run in the browser, but Puter uses browser-direct generation while ComfyUI and Stable Diffusion WebUI/Forge use configurable loopback services. Verify the active connector before processing sensitive media.
What selectable model ID does Timeline Studio provide?
The official source does not publish a selectable model ID for this product. The README names model families and artifacts used by individual features, but it does not define a product-level model ID selector.
Which caption and voiceover systems are documented?
The official README lists Whisper small q8 ONNX for captions; Hojo TTS Light 80M FP16, Kokoro 82M, and browser Piper voices for supported voiceover languages.
What does deterministic offline export guarantee?
The official documentation describes the export as deterministic and offline. It does not state that files exported through every browser, device, and codec environment will be bit-for-bit identical.
Does the MIT License clear every model and generated asset for commercial use?
No. The license notice says third-party models, weights, datasets, fonts, and media retain their upstream terms. Users must also hold the necessary rights to input and published media; face-swap weight commercial rights are not granted or confirmed by the repository MIT License.
Sources and freshness
- Official source
- Official source
- Official source
- Last verified: 2026-08-29
Extended guide
Timeline Studio provides local-first video editing inside a browser while preserving a real, editable multi-track timeline. It is suitable for arranging media, applying supported WebGPU AI operations, producing captions or voiceovers, and completing a deterministic offline export without reducing the project to a one-click generation prompt.
What Timeline Studio includes
As verified on August 29, 2026, the official project describes Timeline Studio as an open-source, local-first browser editor with a CapCut-style multi-track timeline. Local-first is an architectural direction, not a claim that every optional connector is offline. The official source does not publish a selectable model ID for this product.
| Area | Officially documented capability |
|---|---|
| Timeline | Multi-track editing plus source-safe Slip, synchronized Slide, edit-point selection, track locks, source limits, and frame-level rolling adjustments documented on August 27, 2026. |
| AI repair | Browser-local MI-GAN removal, NanoVSR 644K WebGPU 4× restoration, and reversible DRUNet Smart Denoise announced on August 28, 2026. |
| Captions and voices | Whisper small q8 ONNX captions; Hojo TTS Light 80M FP16, Kokoro 82M, and browser Piper voiceovers for the documented languages. |
| Other AI tools | Stable Audio 3 Small Q4 ONNX music, YOLOS tiny detection, MODNet matting, vocal separation, JoyVASA, and LivePortrait workflows. |
| Output | The project explicitly describes deterministic offline export; the documentation does not claim bit-for-bit identical files across every browser, device, or codec environment. |
Recommended editing workflow
- Import and prepare media. Add authorized video, image, and audio assets. Uploaded videos may remain visibly in preparation while metadata and real-presentation-timestamp thumbnails are decoded, as documented on August 28, 2026.
- Build the timeline. Arrange the main visual sequence and supporting tracks. Use trimming, Slip, Slide, edit-point rolling, zoom-aware filmstrips, and track locks while respecting source-media limits.
- Apply AI selectively. Use repair, restoration, denoising, framing, matting, vocal separation, music, or digital-human tools only where needed. Review before-and-after output, especially when an operation changes every frame of a clip.
- Create speech and captions. Generate a supported multilingual voiceover, run automatic captioning with Whisper small q8 ONNX, and review timing, names, mixed-language speech, and recognition cleanup before export.
- Export from the settled timeline. Confirm clip boundaries, audio alignment, captions, repair regions, and the intended output before running deterministic offline export. Preserve the project and source assets if later revision or provenance review may be required.
Privacy and connector boundaries
Browser-local inference can reduce dependence on remote processing, but it should not be treated as a universal no-network guarantee. The August 29, 2026 connector update says Puter uses browser-direct generation, while ComfyUI and Stable Diffusion WebUI/Forge connect to configurable loopback services. Check which feature and connector are active before using confidential footage.
Media-rights checklist
- Use facial media only with the subject’s lawful authorization.
- Do not create illegal, infringing, false, misleading, or non-consensual impersonation content.
- Verify the exact model, checkpoint, dataset, font, stock media, and sample-media terms.
- Do not assume the repository MIT License covers third-party artifacts.
- Treat unclear or conflicting upstream terms as not cleared until the rights holder confirms otherwise.
The Model, Dataset, and Media License Notice states that MIT covers Timeline Studio’s original code and documentation, subject to file-specific notices. It does not automatically license third-party weights or media. In particular, the repository does not confirm commercial rights for the MobileFaceSwap, ArcFace-related, or SCRFD face-swap weights.
Model availability note: The official source does not specify a selectable model ID.
Evidence and freshness
Evidence level: Documentation-verified
AI-assisted editorial content; verify current product details against the linked official sources.
Last verified: