Sourceful

Access Sourceful's Riverflow image editing models through Runware's unified API. State-of-the-art performance for design-grade precision in professional editing workflows.

Introduction

Sourceful is a technology company that provides a full-stack platform for branded packaging design, sourcing, and supply chain management with AI-enabled workflows. Through their platform, brands can generate packaging concepts in seconds, refine designs with human artworkers, access a vetted global supplier network for production, and track sustainability metrics throughout the process.

Sourceful's Riverflow AI models represent a key component of their design capabilities, now available through Runware's unified API. Riverflow achieves state-of-the-art performance in global image editing benchmarks, bringing the same design-grade precision used for print-ready packaging to developers and platforms worldwide. The models excel at complex retouching, object replacement, color correction, lighting adjustment, and text refinement with millimeter-level accuracy required for professional design work.

Image models

Riverflow 1.1 Mini

A fast and cost-efficient image editing model that delivers performance close to Riverflow 1.1 across most tasks, optimized for speed and accessibility.

Model AIR ID: sourceful:1@0.

Supported workflows: Image-to-image editing.

Technical specifications:

  • Positive prompt: Required.
  • Reference images: Supports inputs.references with 1-10 images.
  • Supported dimensions: 1024×1024 (1:1), 1152×864 (4:3), 864×1152 (3:4), 1280×720 (16:9), 720×1280 (9:16), 1248×832 (3:2), 832×1248 (2:3), 1512×648 (21:9), 648×1512 (9:21), 1152×896 (5:4), 896×1152 (4:5).
  • Input image requirements: Width and height between 300-2048 pixels, 20MB file size limit.
{
  "taskType": "imageInference",
  "taskUUID": "f47ac10b-58cc-4372-a567-0e02b2c3d489",
  "model": "sourceful:1@0",
  "positivePrompt": "Remove the coffee stain from the packaging design",
  "width": 1024,
  "height": 1024,
  "inputs": {
    "references": ["c64351d5-4c59-42f7-95e1-eace013eddab"]
  }
}
{
  "taskType": "imageInference",
  "taskUUID": "6ba7b828-9dad-11d1-80b4-00c04fd430c8",
  "model": "sourceful:1@0",
  "positivePrompt": "Change the product color to midnight blue while maintaining all other design elements",
  "width": 1280,
  "height": 720,
  "inputs": {
    "references": [
      "c64351d5-4c59-42f7-95e1-eace013eddab",
      "d7e8f9a0-2b5c-4e7f-a1d3-9c8b7a6e5d4f"
    ]
  }
}

Riverflow 1.1

A versatile daily image editing model supporting up to 10 input images with precise visual edits across 11 aspect ratios and transparent backgrounds. Achieves state-of-the-art performance in global benchmarks.

Model AIR ID: sourceful:1@1.

Supported workflows: Image-to-image editing.

Technical specifications:

  • Positive prompt: Required.
  • Reference images: Supports inputs.references with 1-10 images.
  • Supported dimensions: 1024×1024 (1:1), 1152×864 (4:3), 864×1152 (3:4), 1280×720 (16:9), 720×1280 (9:16), 1248×832 (3:2), 832×1248 (2:3), 1512×648 (21:9), 648×1512 (9:21), 1152×896 (5:4), 896×1152 (4:5).
  • Input image requirements: Width and height between 300-2048 pixels, 20MB file size limit.
{
  "taskType": "imageInference",
  "taskUUID": "f47ac10b-58cc-4372-a567-0e02b2c3d490",
  "model": "sourceful:1@1",
  "positivePrompt": "Change the text from 'ORGANIC COFFEE' to 'PREMIUM BLEND' maintaining the same typography style",
  "width": 1024,
  "height": 1024,
  "inputs": {
    "references": ["c64351d5-4c59-42f7-95e1-eace013eddab"]
  }
}
{
  "taskType": "imageInference",
  "taskUUID": "6ba7b829-9dad-11d1-80b4-00c04fd430c8",
  "model": "sourceful:1@1",
  "positivePrompt": "Replace the glass bottle with an aluminum can while keeping the label design identical",
  "width": 1184,
  "height": 880,
  "inputs": {
    "references": ["c64351d5-4c59-42f7-95e1-eace013eddab"]
  }
}
{
  "taskType": "imageInference",
  "taskUUID": "550e8400-e29b-41d4-a716-446655440010",
  "model": "sourceful:1@1",
  "positivePrompt": "Remove scratches and water drops without altering any other part of the packaging",
  "width": 1280,
  "height": 720,
  "inputs": {
    "references": ["c64351d5-4c59-42f7-95e1-eace013eddab"]
  }
}

Riverflow 1.1 Pro

An advanced image editing model offering superior quality, stability, and accuracy while requiring fewer edit iterations. Designed for professional workflows demanding the highest level of precision.

Model AIR ID: sourceful:1@2.

Supported workflows: Image-to-image editing.

Technical specifications:

  • Positive prompt: Required.
  • Reference images: Supports inputs.references with 1-10 images.
  • Supported dimensions: 1024×1024 (1:1), 1152×864 (4:3), 864×1152 (3:4), 1280×720 (16:9), 720×1280 (9:16), 1248×832 (3:2), 832×1248 (2:3), 1512×648 (21:9), 648×1512 (9:21), 1152×896 (5:4), 896×1152 (4:5).
  • Input image requirements: Width and height between 300-2048 pixels, 20MB file size limit.

This model is currently in early-access status. Contact support for access details.

{
  "taskType": "imageInference",
  "taskUUID": "f47ac10b-58cc-4372-a567-0e02b2c3d491",
  "model": "sourceful:1@2",
  "positivePrompt": "Adjust lighting to create a soft, warm ambiance while preserving all product details and colors",
  "width": 1024,
  "height": 1024,
  "inputs": {
    "references": ["c64351d5-4c59-42f7-95e1-eace013eddab"]
  }
}
{
  "taskType": "imageInference",
  "taskUUID": "6ba7b830-9dad-11d1-80b4-00c04fd430c8",
  "model": "sourceful:1@2",
  "positivePrompt": "Place product inside tissue paper and add a decorative sticker that wraps around both sides as a closure",
  "width": 1280,
  "height": 720,
  "inputs": {
    "references": [
      "c64351d5-4c59-42f7-95e1-eace013eddab",
      "d7e8f9a0-2b5c-4e7f-a1d3-9c8b7a6e5d4f",
      "e8f9a0b1-3c6d-4f8a-b2e4-0d1c2b3a4c5d"
    ]
  }
}
{
  "taskType": "imageInference",
  "taskUUID": "550e8400-e29b-41d4-a716-446655440011",
  "model": "sourceful:1@2",
  "positivePrompt": "Correct color balance and enhance image clarity while maintaining print-ready standards",
  "width": 1024,
  "height": 1024,
  "inputs": {
    "references": ["c64351d5-4c59-42f7-95e1-eace013eddab"]
  }
}