CYAN DOLPHIN PROJECT

Chat / choose a tool

Quick ChatA simple temporary conversation with Qwen or Bonsai. Choose a model, power it on, and talk.
Model ChamberSelect a private local model, configure its context, and work with images or papers.
Model MenagerieExplore the wider local model collection, compare capabilities, and launch one model at a time.
Fiona ChatLaunch a local Fiona who can freely think with, challenge, or precisely recall CLRLY notes.

Image / choose a tool

Creative LabBuild a shared cast, direct shots and plan their images. Visual camera choices, drawing guides and private working drafts.
Image LabCreate with API image models. Search tagged brand photos, build references, explore product angles, and draw edits.
Shared Image ProjectsKeep CLRLY references and finished images together. Share projects, search tags, and build on each other’s work.
Original VISThe familiar Visual Studio: standard images, eight white-background angles, and transformations with GPT Image 2.5.
Layer LabExplore Qwen Image 2.1: native transparency, numbered references, and guided image studies. Research only.
Layer Lab TurboCreate and edit with Qwen Image 2.1 Turbo: fast eight-step renders, numbered references, brand photos, and drawing.
Swap LabTwo images, explicit roles: swap a face or transfer a feature with Qwen Image 2.1. Research only.
Local Image ModelsWake a local image model and adapt the workroom to its verified capabilities.

Video / choose a tool

Creative LabBuild a shared cast, direct shots and plan their images. Visual camera choices, drawing guides and private working drafts.
Local Video · MiniMax H3Wake MiniMax H3 on the RTX 3090 and direct local video from CLRLY frame anchors.

Fiona / choose a tool

Memory StudioCapture, shape, review, and publish the knowledge Fiona is allowed to carry forward.
Fiona WorkshopEdit, test, review, and publish Fiona’s deterministic Telegram responses.

Checking login…

CLRLY / Swap Lab

Opening the workroom…