OpenTryOn v0.0.3 Release Notes
CLI, MCP, and a much wider media model surface
Release Date: 2 August 2026
OpenTryOn v0.0.3 is a large feature release: a unified opentryon CLI, a registry-driven FastMCP server, many new image/video providers (Seedance, Seedream, Kling 3, Luma Ray 3.2, Grok Imagine, Ideogram, Pruna), Kimi understanding models, FASHN / Pruna try-on, and a cleaner separation from the tryon-studio web UI.
What's New
Unified CLI (opentryon)
Installable console script for every registry model:
opentryon <service> --model <model> [params...]
# services: vton | generate | edit | understand | video-generate | bg-removeSupports --dry-run and -o/--output-dir. Local GPU models need pip install opentryon[local].
MCP Server
mcp-server/ is rebuilt on FastMCP 3.x. Tools are generated from tryon.cli.registry (same path as the CLI via invoke_model()). Discovery tools: list_opentryon_tools, opentryon_status.
New media providers
| Area | Models |
|---|---|
| Image generate / edit | Seedream 5.0 Pro, Ideogram 4.0, Grok Imagine Image, Pruna P-Image / Edit / Upscale |
| Video | Seedance 2.5, Luma Ray 3.2, Kling 3.0 / Omni / Turbo, Grok Imagine Video 1.5, Gemini Omni Flash, Pruna P-Video / Replace / Avatar / Animate |
| Try-on | Pruna P-Image-Try-On, FASHN tryon-max / v1.6, Nano Banana 2 Lite (cheap composition path) |
| Understand | Kimi K2.6 / K2.7 Code / K3 (API), Kimi-VL + LLaVA-NeXT (local) |
Demos / product split
The Next.js dashboard prototypes moved to tryon-studio. This package keeps Gradio demos and notebooks only.
Install / Upgrade
pip install -U opentryon
# optional local/GPU models:
pip install -U "opentryon[local]"From source:
git checkout v0.0.3
pip install -e .Docs
- Site: https://tryonlabs.github.io/opentryon/
- Changelog: CHANGELOG.md
- CLI: Unified CLI
- Pruna: API reference
Breaking / migration notes
- Prefer
opentryon/ MCP over ad-hoc root scripts (legacy examples live underexamples/legacy/andtests/legacy/). - Web UI consumers should point at tryon-studio + MCP, not removed
demo/virtual-tryonpaths. - Core
pip install opentryonstays cloud-light; torch/transformers are only in the[local]extra.