Pricing
The homepage says Floyo is free to start and that users pay only when they generate. Teams get pooled GPU credits, and Enterprise adds dedicated GPU pools. Specific prices are on the pricing page.
Learn moreRun ComfyUI and open-source AI workflows in the browser with team controls

Floyo is a browser platform for running ComfyUI and other open-source AI workflows on hosted GPUs. It combines a large model and node library with shared workspaces, run history, model governance, an agent and an API for studios and creative teams.
Floyo is a browser-based platform for running AI generation workflows, built around the ComfyUI ecosystem. Instead of installing models and nodes on a local machine, users open a workflow in the browser, supply their own inputs and run it on hosted GPUs. The catalog covers open and closed models from providers such as Flux, Wan, Qwen, Kling, Luma, OpenAI and Z-Image, with more than 850 models and over 2,500 vetted nodes listed on the homepage. Users can also upload workflows they built in ComfyUI, or bring their own LoRAs, fine-tunes and custom nodes. The product is aimed at creative professionals, studios, agencies and universities, plus developers who want generation behind an API. Solo users can start with ready-made templates: open a workflow, read the instructions, change the input and press Run, without touching the node graph. Teams get shared workspaces, pooled GPU credits, role-based access and a run history that logs the settings, inputs and outputs of every generation, so a colleague's earlier run can be reopened, adjusted and rerun. Floyo Agent, currently in beta, takes a plain-language request, searches workflows, models and nodes, and chains multi-step tasks. Every workflow is also exposed as an API endpoint and through an MCP server for external agents. Governance is the main differentiator from running plain ComfyUI. Models are tagged as Verified Private (a contract bars the provider from training on user content) and as cleared for commercial use. Administrators can set policy per team, auto-block models that fail either test, block provider families or block individual nodes, including community-sourced ones, with enforcement on both UI and API runs. The company says its security posture has passed Amazon MGM Studios and Netflix review, and an Enterprise tier adds dedicated GPU pools and commercial-use indemnification. Among alternatives, Floyo sits between self-hosted ComfyUI, which offers freedom without oversight, and single-model generators that limit users to one route. It suits teams that want open-source flexibility with production-grade controls.




The homepage says Floyo is free to start and that users pay only when they generate. Teams get pooled GPU credits, and Enterprise adds dedicated GPU pools. Specific prices are on the pricing page.
Learn moreThe homepage points to documentation for Floyo Agent and run history, a Trust Center for security and compliance details, and a Book a Demo option for teams. Enterprise walkthroughs are published for sharing with teams.
Learn moreEvery workflow is a live API endpoint and is accessible via an MCP server for external agents. Workflows built in ComfyUI can be uploaded, and models from providers such as OpenAI, Google, Kling, Luma and Flux can be called inside one workflow.
Learn moreBrowser-based ComfyUI workflows, 850+ open and closed models, 2,500+ vetted nodes, custom LoRAs and nodes, Floyo Agent (beta), shared workspaces, role-based access, run history, per-team model and node policy, Verified Private and commercial-use model tags, API and MCP access.
Learn more




