dMaya

Use it when

You want vibe-style product UI on a multi-screen canvas—iterate with an agent, share preview links for approval, then export HTML/CSS/Tailwind for engineering or MCP-connected coding agents.

What it does well

• Prompt-to-canvas workflow with desktop/mobile frames side by side, not buried in a chat thread
• Multi-model picker (Gemini, Kimi, Claude, GPT, and more) with per-model credit multipliers; agent can choose when you do not
• Reusable components and design tokens that stay in sync across screens; plan/research modes for harder briefs
• Shareable preview links with device toggle and visual comments—clients can review without an account
• Export approved screens as HTML/CSS/Tailwind for handoff; every plan includes unlimited projects and unlimited team members

Skip it when

• You expect free generation after signup—observed accounts start at No plan with 0 credits until you buy a tier
• Starter’s 3,000 credits/mo (about 6 per chat message or 22 per screen on the default model) will not cover steady client volume
• You need annual credit rollover—that sits on Pro/Scale yearly, while Starter credits expire monthly
• You want chat-to-full-stack app shipping (Softgen/Lovable) or live React-repo editing (Onlook/Subframe)
• Heavy use of high-multiplier models would burn the same pool much faster than the default-model estimates

Fits your workflow

Describe or attach references → generate screens on the canvas → refine via chat/hand edits → share preview for feedback → export HTML/CSS/Tailwind to engineering or agents. Pick Starter (3,000 credits/mo) to explore; Pro (10,000) for steady work; Scale (25,000) for high volume. Top up 1,000-credit packs when a sprint spikes. Finish production logic in your repo.