Overview

Runs Topaz upscaling on a hosted image. Supports 1–4× scale (2× and 4× are most useful) with five Topaz presets. image_url is required — fabricated URLs are caught by HEAD preflight before fal is called.

Auth

oauth-jwt

Returns

{ imageUrl, generationId, creditsCharged, creditsRemaining, lowCreditsWarning? }

Web client

src/app/[locale]/dashboard/creative-studio/tools/upscale-image/upscale-image-client.tsx

MCP handler

mcp-director/src/tools/studio.py

Quick start

Invoke the tool over Streamable HTTP at {MCP_BASE_URL}/mcp/. The request below assumes you already minted an MCP JWT via the OAuth flow.

MCPcurlTypeScript

Copy

{
  "name": "studio_upscale_image",
  "arguments": {
    "image_url": "<image_url>"
  }
}

Parameters

Name Type Default Description
image_urlrequired string — Public URL of the source image. Fabricated URLs rejected.
upscale_factor 1 | 2 | 3 | 4 2 Upscale factor.
topaz_model "Low Resolution V2" | "Standard V2" | "High Fidelity V2" | "Text Refine" | "CGI" "Standard V2" Topaz preset (distinct from the fal app id passed in model).
face_enhancement boolean true Enable Topaz's face refinement pass.
output_format "jpeg" | "png" "jpeg" Output image format.
model string "fal-ai/topaz/upscale/image" fal app id.

Examples

4× upscale with High Fidelity

json

Copy

{
  "name": "studio_upscale_image",
  "arguments": {
    "image_url": "https://v3b.fal.media/files/.../portrait.jpg",
    "upscale_factor": 4,
    "topaz_model": "High Fidelity V2",
    "face_enhancement": true
  }
}

Performance