veida-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@veida-mcp Generate a 16:9 image of a lighthouse on a rocky cliff during a storm."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Veida MCP
An MCP server that actually generates images, with no API key and no account.
Most image MCP servers are a wrapper around a key you have to go and buy first. This one talks to the anonymous tier of Veida — a client id in a header instead of a login — so it works the moment it is installed.
> generate a 16:9 cover image of a paper boat on a calm lake at sunrise
→ https://cdn.veida.ai/uploads/kie/image/….webpTools
Tool | What it does |
| Prompt → hosted image URL. Submits, polls, returns. 40–120 seconds. |
| The model shelf, which ones run signed-out, and the page for each. |
| Pages worth sending a person to — editor, upscaler, background remover. |
| The exact ceilings, so an agent does not build something that always fails. |
Related MCP server: Replicate MCP Server
Install
Claude Desktop / Claude Code
{
"mcpServers": {
"veida": { "command": "npx", "args": ["-y", "veida-mcp"] }
}
}Cursor / Cline / Windsurf
Same block, in that client's MCP config file. There is nothing else to set — no
env, no key.
What the free tier is, measured
The anonymous grant is 4 credits and one image costs 4 — so the server mints a fresh client id per image rather than reusing one.
A ceiling of 30 credits per IP per day sits on top: roughly 7 images a day from one machine, however many ids it invents.
Free output is 1K and watermarked.
🔴 Editing an existing image cannot run anonymously. The editing model costs 30 credits against a 4-credit grant, so an anonymous edit can never be paid for. The tool set says so rather than offering a call that always fails; for edits, sign in and use the image editor or the ChatGPT image editor.
There is no video on this site, so no video tool.
Signing in removes the watermark, lifts the cap and opens the rest of the shelf. The current lineup is on the models the generator offers: GPT Image 2 when the picture has to contain readable text, Nano Banana 2 for instruction following, Seedream 5.0 Lite for colour and composition, and Qwen Image 3 for typography. Pricing lists what each tier costs.
Writing prompts that work
Name the light, the material and the composition. "A product photo of a mug" gives the model nothing; "matte black ceramic mug on pale oak, soft window light from the left, shallow depth of field" gives it a picture. Worked examples live in the prompt library, and the showcase is there if you want to judge output quality before wiring anything up.
The same engine is what the ChatGPT image generator page serves in a browser — this server is the headless door to it.
Develop
npm install
npm run build
node dist/index.js # speaks MCP over stdioLinks
MIT.
Available Tools
4 toolsbrowse_toolsA
List the Veida pages worth sending a person to — generator, editor, upscaler, background remover, prompt library, pricing.
Inputs: no named parameters.
free_tier_limitsA
Explain exactly what the anonymous free tier allows, so the caller does not build something that always fails.
Inputs: no named parameters.
generate_imageA
Generate an image from a text prompt and return a hosted image URL. No API key or account is needed — this runs on the anonymous free tier. Output is 1K and watermarked. Takes 40-120 seconds.
Inputs: prompt (required), aspectRatio, timeoutSecs.
list_modelsA
List the image models Veida serves, which of them a signed-out caller can run, and the page a human can run each one on.
Inputs: scene.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 4 tools
generate_image is the only action tool, while list_models, browse_tools, and free_tier_limits are distinct informational resources. There is minor overlap between list_models and browse_tools in that both mention pages a human can visit, but their primary intents differ enough.
Three tools use a clear verb_noun pattern (generate_image, list_models, browse_tools), but free_tier_limits is a noun phrase rather than get_free_tier_limits or explain_free_tier_limits, creating a small inconsistency.
Four tools is a tight, appropriate scope for an anonymous free-tier image generation service. Each tool has a clear purpose and none feel redundant.
The set covers the main programmatic action (generate_image) plus the essential supporting information a caller needs: available models, human pages, and free-tier limits. A model-selection parameter on generate_image or a separate generate_with_model tool would close the most obvious minor gap.
Maintenance
Related MCP Connectors
Image and video AI tools and your own pipelines, run from any AI assistant.
LLM chat, text tools, image generation, editing, batch image jobs, and asynchronous video generation
Generate AI images, video, voiceovers and music from Claude, ChatGPT or Cursor through 50+ models (Veo 3.1, Kling 3, Seedance, Nano Banana, GPT Image, ElevenLabs). Also image editing, upscaling, background removal, face swap, transcription, voice cloning and UGC-style video ads. Sign in with OAuth — no API key to paste. Tools are annotated (read-only vs. credit-spending); failed generations are refunded.
Generate images, videos and music with model discovery, cost estimates and asynchronous results.
Related MCP Servers
- AlicenseDqualityDmaintenanceEnables image generation, editing, and description using Ideogram AI's models through natural language.44MIT
- AlicenseNot gradedqualityDmaintenanceEnables AI image generation and editing using Replicate's official models like Flux, SDXL, and Seedream, with tools to search models and generate images.6MIT
- AlicenseAqualityDmaintenanceEnables AI image and video generation using Google Nano Banana and Veo 3.1 via a LiteLLM gateway, providing tools for synchronous image generation and asynchronous video generation with polling, returning public URLs.4MIT
- FlicenseNot gradedqualityAmaintenanceEnables AI agents to generate images from text prompts, list available models, and view generation history through stdio, SSE, or Streamable HTTP transports.-