Image and PDF rendering for your AI agent

Give your agent a rendering tool: it supplies a structured template, text, images and data; Rendry composes the finished file. Reuse the layout for product cards, certificates, invoices and other documents, changing the data for each output.

The MCP server exposes 33 tools to an external client such as Claude, Cursor or Codex. That client can write copy and design layouts before calling Rendry. Rendry handles template validation, previews and rendering; the rendering pipeline does not run a generative image model. You can open the same template in the visual editor and adjust it.

Start free; 50 credits included · Documentation

Why teams and agents use it

  • Reusable layouts. Fonts, colours, image positions and variable fields live in the template. Send different data to render another version of the same design.
  • MCP integration. Connect your external agent to https://rendry.io/api/v1/mcp to create templates from JSON DSL, preview drafts and render images or documents.
  • Manual editing. Agent-created templates open in the visual editor. Review the layout, adjust layers and save the version your automation will use.
  • Prepaid usage. 50 free credits to start; one-time packs from $15, no subscription. Credits pay for rendering and supported image-processing operations.

Ready-made templates for this

  • YouTube Clickbait Thumbnail: YouTube 1280x720 clickbait. Photo background, huge white stroked words + yellow money line, red arrow, corner badge. Mood: maximum urgency. Use for: challenges, experiments, vlogs. (1280x720)
  • Subject Glow Poster: Poster with a background-removed subject popping over a designed glow background. Mood: editorial depth. Use for: hero posters from a photo. (1080x1350)
  • Glassmorphism Card: Frosted-glass card over a colorful background. Mood: trendy UI. Use for: feature highlights, covers. (1200x630)
  • Sale Banner: 1200x628 sale banner. Hot red-orange mesh + grain, giant gradient percent, tilted yellow badge, white CTA. Mood: loud urgency. Use for: sales, promos, flash deals. (1200x628)
  • Product Card: 1080x1350 product card. Light pastel mesh, product photo with soft shadow, red circle discount badge, gradient price, dark CTA pill. Mood: clean marketplace. Use for: e-commerce listings, catalogs. (1080x1350)
  • Event Poster: 1080x1350 night event poster. Neon violet-magenta-cyan mesh + heavy grain, glowing gradient title, line-up list, date chip. Mood: rave neon. Use for: parties, concerts, club nights. (1080x1350)

Connect your AI agent through MCP

Rendry exposes template design and rendering through a remote MCP server (https://rendry.io/api/v1/mcp) with 33 tools. Your external AI client can supply layouts, copy and data, preview a template, then request finished images or documents. Rendry renders those structured inputs; it does not run a generative image model in the rendering pipeline. The same template opens in the visual editor for manual adjustments.

Machine-readable: llms.txt · llms-full.txt · openapi.json · mcp.json

Frequently asked questions

Does Rendry generate artwork from a text prompt?
The rendering pipeline takes structured templates and data, not a text-to-image prompt. An external AI client can supply copy and layouts through MCP. Separate image-processing tools use ML: background removal processes an existing image, and optional object extraction uses SAM3 through fal.ai to segment an object by description or point.
How does an AI agent use Rendry?
Connect the remote MCP server (https://rendry.io/api/v1/mcp) with one URL; claude.ai connects via OAuth, CLI agents use an API key. The agent then lists templates, designs its own from JSON DSL, previews drafts and renders finals.
Can I fix what the agent designed?
Yes. Agent-created templates are normal Rendry presets; open them in the visual drag-and-drop editor, move layers, change fonts, save. The agent can keep rendering the adjusted version.
What does it cost?
Every new account gets 50 free credits with full API and MCP access, no credit card. A final image costs 1-3 credits by size, a preview 0.1. A standard PDF or HTML document costs 2 credits for up to 20 pages; each further started block of 20 pages adds 1 credit. Large pages flattened entirely to pixels can add a compute surcharge; a free dry run quotes the exact cost. One-time packs: $15 for 500, up to $199 for 100,000.

Related