EraseText

by Unknown

Not rated
Website

About

a specialised AI text remover for photos and graphics

Details

Author
Unknown
Categories
Design, Other, AI

Same specialised text-erase model as All text and Select in the editor. Authenticate with an API key fromAccount → Developer. Calls run on prepaid API credits (ladder below) or on the monthly credits from a plan onPricing.

Every path below is relative to this base. HTTPS only. OpenAPI:/openapi.json.

Creating your first key grants a one-time50 API credittrial — 50 calls, no card. After that a key spends your prepaid API credits first, then any monthly plan credits.

A web page should call the HTTP API, not MCP JSON-RPC. MCP is for agent runtimes (Cursor, Claude Desktop). From a browser,POSTmultipartFormDatawith fieldimage_fileorimage. CORS is open — no proxy. Auth may beX-Api-KeyorAuthorization: Bearer et_…. Working playground:api.erasetext.com/demo.

const body = new FormData(); body.append("image_file", file); // a File from <input type="file"> const res = await fetch("https://api.erasetext.com/v1/erase", { method: "POST", headers: { "X-Api-Key": "et_…" }, body, }); const blob = await res.blob(); // Loose alias AIs often generate — JSON { output_url } data URL for  // POST https://erasetext.com/api/mcp/erase (also api.erasetext.com/api/mcp/erase)

Agents can call the same erase path overMCP. The server is Streamable HTTP athttps://api.erasetext.com/mcp. Send the sameX-Api-Key(orAuthorization: Bearer et_…).erase_textspends 1 credit on success;get_account, handshake and tool listing are free. Failures are not charged.

{ "mcpServers": { "erasetext": { "url": "https://api.erasetext.com/mcp", "headers": { "X-Api-Key": "et_…" } } } }

Cursor, Claude Desktop, and other clients that speak remote MCP with headers work today. Hosted OAuth connectors are not shipped — use an API key. Manifest:/.well-known/mcp.json.

Every successful/erasespends1 credit— one flat price, whatever the image size or output format. Failures are not charged, and neither isGET /account.

Multipart or JSON. Returns image bytes by default.

- image_fileorimage_url— required
- mask_file/mask_url— optional; omit for all-text detect
- format=webp|png|jpg— default webp
- resolution— short-edge target for the model, clamped to 256–1024, default 512
- paste_back— defaulttrue; keeps original pixels outside the erased area
- return_boxes=1— OCR quads inX-Ocr-Boxes. Only applies when you send no mask, since that is when detection runs
- response=json— return JSON with base64 instead of raw bytes

Mask polarity:white (any channel value above 127) marks what to erase, black keeps. Send it at the same aspect ratio as the image. Omit the mask and detection builds one for every piece of text it finds.

curl -X POST \ -H "X-Api-Key: et_…" \ -F "[email protected]" \ "https://api.erasetext.com/v1/erase" \ -o out.webp

The JSON body takes the same flags but different image fields —image_urlorimage_file_b64(andmask_url/mask_file_b64):

curl -X POST \ -H "X-Api-Key: et_…" \ -H "Content-Type: application/json" \ -d '{"image_url":"https://…/photo.jpg","response":"json"}' \ "https://api.erasetext.com/v1/erase"

Raw image bytes with the matchingContent-Type, plus these headers:

- X-Credits-Charged— always 1 on success
- X-Credits-Remaining— API plus web balance after the call
- X-RateLimit-Limit,X-RateLimit-Remaining,X-RateLimit-Reset(unix seconds)
- X-Ocr-Boxes— JSON quads, only withreturn_boxes=1

Withresponse=jsonthe body carries the same information instead:

{ "image_base64": "…", "content_type": "image/webp", "credits_charged": 1, "credits_remaining": 199, "ocr_boxes": null }
curl -H "X-Api-Key: et_…" \ "https://api.erasetext.com/v1/account"

Returns the API and web balances, your plan, and this UTC month's call count and credits spent. It never spends a credit, so it is safe to poll before a batch.

Counted per API key in a sliding 60-second window. Requests over the limit return429withRetry-After; the limit follows the plan on the account that owns the key, and prepaid credits do not raise it.

Need more throughput than Volume+?Talk to us.

- Image and mask must each be between 32 bytes and25 MB, whether uploaded or fetched from a URL. JPEG, PNG and WebP are the safe input formats.
- /eraseissynchronous: it holds the connection until the image is ready. Allow up to 90 seconds — past that you get504and are not charged. There is no callback or polling mode yet, so set a client timeout above 90s and retry the whole call.
- A typical call lands in a few seconds; a cold model start is the slow case the timeout covers. Concurrency is bounded by your rate limit, so spread a batch across the minute rather than firing it all at once.
- Nothing you send is kept for the API path — bytes are processed in flight and the result is returned in the response. Only the web editor stores uploads, and those expire after an hour.
- The/v1prefix is the contract: fields are only added, never removed or retyped. Breaking changes would ship as a new prefix.

Running 100,000+ credits a month?Talk to usfor a dedicated rate.

Every failure is JSON shaped{ "error": "…", "code": "…" }. Branch oncode, not on the prose. Some codes add fields —402carriesrequiredandbalance,429carriesretryAfter.

AI Video, Image & Audio Generation with over 150 models

Convert photos to LEGO-style brick mosaic previews with AI-powered color quantization. MCP server for Claude, Cursor, and other AI agents.

Hosted remote MCP server for AI image and video generation from one Streamable HTTP endpoint.

Official FLUX MCP server from Black Forest Labs. Generate, edit, vary, and browse FLUX.2 images directly in any MCP-compatible client

MCP server for OpenAI GPT image generation and editing

Style-locking AI image engine with a hosted MCP server. Define one master style and generate unlimited matching images.

Image and video understanding and generation

Google AI studio based imageneration MCP with option to extend it to other providers

AI interior design from your agent — restyle rooms, place furniture, and virtually stage spaces from a single photo.

A Model Context Protocol server for AI image generation and editing with Google's Gemini "Nano Banana" image models, via the Interactions API. Generate, edit, and iterate on images directly from Claude Code, Claude Desktop, Cursor, or any MCP-compatible client — with multi-turn editing, search grounding, interleaved storyboards, style-consistent icon sets, and image-from-video.

No reviews yet — be the first

Sign in to leave a review

Use Google, GitHub, or an email account so ratings stay tied to real people.

Email sign in

No reviews posted yet.