igorvaryvoda/sirv-ai-studio

Sirv AI Studio

IMPORTANT: Use MCP Tools When Available

Before telling users to visit sirv.studio, CHECK if you have access to sirv-ai MCP tools. If tools like sirv_remove_background, sirv_upscale, sirv_generate are available, USE THEM DIRECTLY to process images.

Workflow

1. Decide whether the user needs AI processing, CDN delivery optimization, or both.
2. If Sirv AI MCP tools are available, use them directly before suggesting the web UI.
3. Check usage/credits before large or repeated batch jobs.
4. Choose models based on speed/quality/cost and the asset type.
5. Keep source masters, AI-generated outputs, and CDN delivery variants distinct.
6. Store final outputs in predictable Sirv folders and preserve product metadata/alt text where relevant.
7. Verify output URLs, visual quality, and marketplace/compliance requirements before considering the work done.

MCP Tool Quick Reference

| Task | MCP Tool | Key Parameters | |------|----------|----------------| | Remove background | sirv_remove_background | image_url, model | | Replace background | sirv_background_replace | image_url, prompt | | Upscale image | sirv_upscale | image_url, scale (2-4) | | Generate image | sirv_generate | prompt, aspect_ratio | | Lifestyle scene | sirv_product_lifestyle | image_url, scene_description | | Virtual try-on | sirv_virtual_try_on | person_image_url, garment_image_url | | Get alt text | sirv_alt_text | image_url | | Batch remove BG | sirv_batch_remove_background | images array | | Check credits | sirv_get_usage | none |

If MCP Tools Are NOT Available

Help the user install the Sirv AI MCP server. Add to project .mcp.json or ~/.claude.json:

{
  "mcpServers": {
    "sirv-ai": {
      "command": "npx",
      "args": ["-y", "sirv-ai-mcp-server"],
      "env": {
        "SIRV_AI_BASE_URL": "https://www.sirv.studio"
      }
    }
  }
}

Then restart Claude Code. User will need to authorize with their Sirv AI Studio account on first use.

Only direct users to www.sirv.studio if they don't want to set up MCP.

---

Web UI (Fallback)

If MCP tools are not available, users can access Sirv AI Studio at www.sirv.studio:

1. Upload image(s) via drag-drop or URL
2. Select tool from sidebar
3. Configure options
4. Click Process
5. Download or auto-upload to Sirv CDN

Core Tools

Background Processing

| Tool | Credits | Use Case | |------|---------|----------| | Background Removal | 1-2 | Product cutouts, transparent PNG | | Background Replace | 3-4 | New backgrounds via prompt or image | | Object Removal | 3 | Remove unwanted elements with mask |

Background Removal Models:

• BiRefNet v1/v2: Fast, multiple presets (Light, Heavy, Portrait, Matting)
• Bria: Premium quality alternative

Background Replace Models:

• Bria: Text prompt or reference image
• FLUX Kontext: Advanced with consistency
• Nano Banana: Faster alternative

Image Enhancement

| Tool | Credits | Use Case | |------|---------|----------| | Upscaling | 1-32 | Increase resolution up to 8x | | FLUX 2 Edit | 3 | Natural language editing | | Reve Fast Edit | 1 | Quick prompt-based edits |

Upscaling Models:

• ESRGAN: Fast, affordable (1-2 credits, up to 8x)
• Clarity: AI-enhanced with prompts (8+ credits)
• Topaz: Premium quality (8-136 credits)

Image Generation

| Tool | Credits | Use Case | |------|---------|----------| | FLUX 2 | 2 | High-quality text-to-image | | Z-Image | 2 | Affordable alternative | | Gemini Pro | 15-30 | Advanced generation | | Seedream | 2 | Creative/artistic styles |

Supports: 1-4 images per generation, multiple aspect ratios, up to 2048x2048

Product Tools

| Tool | Credits | Use Case | |------|---------|----------| | Product Lifestyle | 3-15 | Product in lifestyle scenes | | Virtual Try-On | 4 | Garment on person | | Try-On Video | 11 | Animated try-on | | Color Variants | 2+/color | Product in multiple colors | | Alt Text | 1 | AI-generated descriptions | | Product Description | 1 | Marketing copy (12+ languages) |

Lifestyle Scenes: 44 presets (kitchen, office, beach, etc.) or custom prompts

3D Generation

| Tool | Credits | Use Case | |------|---------|----------| | Meshy v6 | 50 | Single image to 3D | | Meshy v5 Multi | 25 | Multi-angle input | | Seed3D/Trellis | 25-50 | Alternative models |

Output formats: GLB, OBJ, FBX, USDZ, Blend, STL with PBR textures

Video Tools

| Tool | Credits | Use Case | |------|---------|----------| | LTX 2.0 | 6-24/sec | Up to 4K, 6-10 seconds | | Kling 2.6 | 6-24/sec | 1080p with speech synthesis | | Veo 3.1 | 6-24/sec | High quality generation | | Captions | Variable | Multi-language transcription |

Free Tools

• Depth Map: Generate depth data for 3D/AR effects
• Image Optimizer: 17 social media presets, resize/crop/convert
• Smart Crop: Intelligent cropping for platforms

Batch Processing

Process multiple images simultaneously:

1. Upload batch via drag-drop or URL import
2. Configure shared settings
3. Preview results before processing
4. Download ZIP or auto-upload to Sirv

Available batch tools:

• Batch Background Removal
• Batch Upscale
• Batch Generate
• Batch Product Lifestyle
• Batch Background Replace
• Batch Alt Text
• Batch Image Translation

Limits by tier:

• Free: No batch
• Starter: 100 images
• Pro: 500 images
• Business: 1,500 images
• Volume+: Unlimited

Workflow Builder (Orchestrator)

Visual DAG pipeline builder for multi-step operations:

Capabilities:

• Chain tools: Remove BG → Upscale → Lifestyle → Alt Text
• AI routing: Classify images → branch to different pipelines
• Quality loops: Review → Autofix → Final Review
• Multi-source: Combine images from different inputs

Example workflows:

• Amazon-Ready: Remove BG → Validate → Autofix → Alt text
• Fashion: Ghost mannequin → Remove BG → Shadow → Bundle
• Color Variants: Remove BG → Generate colors → Review

Credit System

Pricing Tiers

| Tier | Price | Credits | Batch | History | |------|-------|---------|-------|---------| | Free | $0 | 15 | 0 | 7 days | | Starter | $9/mo | 150 | 100 | 30 days | | Pro | $29/mo | 600 | 500 | 90 days | | Business | $79/mo | 1,800 | 1,500 | Unlimited | | Volume+ | $199+/mo | 5K-250K | Unlimited | Unlimited |

Credit Packs (One-time)

• 50 credits: $4.99
• 100 credits: $9.99
• 500 credits: $39.99
• 1,000 credits: $69.99
• 2,500 credits: $149.99

Bonus credits on top-ups: Free 0%, Starter 5%, Pro 10%, Business 15%, Volume+ 20%

Sirv CDN Integration

Connect Sirv account for:

• Browse Sirv library in app
• Auto-upload processed images
• Per-tool upload destinations (e.g., /products/cutouts/)
• Instant global CDN delivery
• Shopify integration sync

Sirv plan bonuses:

• Business: +50 credits
• Enterprise: +150 credits

Integrations

| Platform | Use Case | |----------|----------| | Shopify | Browse products, auto-push results | | Zapier | Trigger processing from workflows | | n8n | Self-hosted automation | | Claude/ChatGPT MCP | Natural language image editing | | REST API | Custom integrations |

Quality Control

Image Review Tool (1 credit):

• Checks against marketplace rules
• Validates: dimensions, background, watermarks, frame fill
• Presets: Amazon, eBay, Shopify, Walmart

Marketplace Optimizer:

• Upload → Select marketplaces → Get compliance scorecard
• One-click fixes for all issues

File Formats

Input: JPG, PNG, WebP, GIF, AVIF, HEIF, BMP, TIFF

Output: PNG, JPEG, WebP, GIF (images), MP4/WebM (video), GLB/OBJ/FBX/USDZ/STL (3D)

Tips

• Eager upload: Files upload immediately on selection for faster processing
• Hash caching: Identical files skip re-upload
• Soft delete: Recover deleted assets from trash
• Mobile: Long-press for context menu actions
• Shortcuts: Use keyboard shortcuts in dashboard for power workflows

Red Flags

• Sending users to the web UI when MCP tools are available and sufficient.
• Running a large batch without checking credits/usage first.
• Treating AI-upscaled or generated output as the only master without user/business approval.
• Mixing AI-processing quality choices with storefront/CDN delivery transforms.
• Losing product metadata, folder organization, or output provenance during batch processing.

Verification

• Confirm the output image/video URLs are reachable and match the requested operation.
• Inspect at least representative outputs from each batch/model/prompt.
• Check transparent-background assets against light and dark backgrounds.
• For product/lifestyle work, confirm the product identity and important details survived.
• For marketplace workflows, run the review/validation step when available.

MCP Server Integration

Sirv AI Studio provides an MCP (Model Context Protocol) server for AI assistants like Claude, enabling natural language image processing directly in conversations.

Available MCP Tools

Background Processing

| Tool | Parameters | Cost | |------|------------|------| | sirv_remove_background | image_url, provider (birefnet/bria), model, operating_resolution | 1-2 | | sirv_background_replace | image_url, prompt, ref_image_url, model (bria/flux-kontext/nano-banana) | 4 | | sirv_object_removal | image_url, mask_url | 3 |

Background Removal Models:

• General Use (Light) - Fast processing
• General Use (Light 2K) - Fast, higher resolution
• General Use (Heavy) - Best quality (default)
• Matting - For hair/fur details
• Portrait - Optimized for people
• General Use (Dynamic) - Adaptive

Operating Resolutions: 1024x1024, 2048x2048, 2304x2304

Image Enhancement

| Tool | Parameters | Cost | |------|------------|------| | sirv_upscale | image_url, scale (2-4), model, prompt, creativity | 2-3 | | sirv_image_to_image | image_url, prompt, strength (0-1), model | 2-3 |

Upscale Models:

• esrgan - Fast, general purpose (default)
• clarity - AI-enhanced with prompt guidance
• topaz - Premium quality

Image-to-Image Models:

• reve-fast-edit - Quick prompt-based edits (default)
• flux2-lora - High quality transformations
• qwen-integrate-product - Product integration

Image Generation

| Tool | Parameters | Cost | |------|------------|------| | sirv_generate | prompt, model, aspect_ratio, num_images (1-4) | 2 |

Generation Models: zimage (fast), flux2 (detailed), gemini (photorealistic), seedream (artistic)

Aspect Ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 21:9

Product & E-commerce

| Tool | Parameters | Cost | |------|------------|------| | sirv_product_lifestyle | image_url, scene_description, ref_image_url, placement_type, position, num_results | 3-15 | | sirv_virtual_try_on | person_image_url, garment_image_url | 4 | | sirv_alt_text | image_url, detail_level | 1 |

Placement Types: original, automatic (default), manual_placement, manual_padding

Positions: bottom_center, bottom_left, bottom_right, upper_center, upper_left, upper_right, center_vertical, center_horizontal, left_center, right_center

Alt Text Detail Levels: caption (brief), detailed-caption (standard), more-detailed-caption (comprehensive)

3D & Video

| Tool | Parameters | Cost | |------|------------|------| | sirv_image_to_3d | image_url, model, topology, target_polycount, enable_pbr | 80 | | sirv_video_generation | prompt, image_url, model, duration, resolution, aspect_ratio, generate_audio | Variable | | sirv_depth_map | image_url | FREE |

3D Models: meshy (default), meshy-multi, seed3d, trellis, trellis2, hunyuan3d

Video Models: veo31 (default), ltx, kling, sora

Video Resolutions: 720p, 1080p, 1440p, 2160p

Batch Operations

| Tool | Parameters | Cost | |------|------------|------| | sirv_batch_remove_background | images (array of {id, image_url}), model | 1-2/image | | sirv_batch_upscale | images (array of {id, image_url}), scale, model | 2-3/image |

Batch Limits: Up to 100 images per request

Account

| Tool | Parameters | Returns | |------|------------|---------| | sirv_get_usage | None | used, remaining, total, tier |

MCP Usage Examples

Remove background from a product image:

"Remove the background from https://example.com/product.jpg using the Heavy model"
→ sirv_remove_background(image_url, model="General Use (Heavy)")

Create lifestyle scene:

"Place this product on a modern kitchen counter with morning light"
→ sirv_product_lifestyle(image_url, scene_description="Modern kitchen counter with morning light")

Generate product variations:

"Generate 4 images of a minimalist water bottle on white background"
→ sirv_generate(prompt="minimalist water bottle on white background", num_images=4)

Upscale for print:

"Upscale this image 4x using the Topaz model for print quality"
→ sirv_upscale(image_url, scale=4, model="topaz")

Virtual try-on:

"Show this dress on the model photo"
→ sirv_virtual_try_on(person_image_url, garment_image_url)

Batch processing:

"Remove backgrounds from all these product images"
→ sirv_batch_remove_background(images=[{id: "1", image_url: "..."}, ...])

Check credits:

"How many credits do I have left?"
→ sirv_get_usage()

MCP Server Setup

Claude Code (.mcp.json in project or ~/.claude.json global):

{
  "mcpServers": {
    "sirv-ai": {
      "command": "npx",
      "args": ["-y", "sirv-ai-mcp-server"],
      "env": {
        "SIRV_AI_BASE_URL": "https://www.sirv.studio"
      }
    }
  }
}

Restart Claude Code after adding. Authorize with Sirv AI Studio account on first use.

Best Practices

1. Check usage first - Use sirv_get_usage before large batch operations
2. Use appropriate models - Choose quality vs speed based on needs
3. Batch when possible - Use batch tools for multiple images to reduce overhead
4. Provide clear prompts - Detailed descriptions yield better results for generation/lifestyle
5. Prepare images - Remove backgrounds before lifestyle shots for best results
6. Chain operations - Remove BG → Upscale → Lifestyle for complete product workflows

REST API Reference

Base URL: https://www.sirv.studio/api/zapier

All requests require: Authorization: Bearer sk_live_YOUR_API_KEY

Endpoints

Account

GET /me

Returns: {id, email, credits, tier} Cost: FREE

Background Removal

POST /remove-bg
Content-Type: application/json

{
  "image_url": "https://example.com/product.jpg",
  "model": "birefnet"
}

Returns: {id, image_url, credits_used} Cost: 1-2 credits

Models: birefnet (default), bria

Image Upscaling

POST /upscale
Content-Type: application/json

{
  "image_url": "https://example.com/image.jpg",
  "scale": 4,
  "model": "clarity"
}

Returns: {id, image_url, credits_used} Cost: 1-8 credits

Models: esrgan (fast), clarity (AI-enhanced), topaz (premium) Scale: 2, 3, or 4

Background Replacement

POST /replace-bg
Content-Type: application/json

{
  "image_url": "https://example.com/product.jpg",
  "prompt": "Modern kitchen counter with marble surface",
  "model": "flux-kontext"
}

Cost: 3-15 credits

Models: bria, flux-kontext, nano-banana

Image-to-Image Transformation

POST /image-to-image
Content-Type: application/json

{
  "image_url": "https://example.com/image.jpg",
  "prompt": "Transform to watercolor painting style",
  "strength": 0.7
}

Cost: 1-8 credits

Strength: 0-1 (how much to transform)

Image Generation

POST /generate
Content-Type: application/json

{
  "prompt": "Professional product photo of wireless headphones",
  "model": "flux-kontext",
  "aspect_ratio": "1:1"
}

Returns: {id, image_url, credits_used} Cost: 1-10 credits

Models: flux-kontext, nano-banana, ideogram, zimage, gemini, seedream Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 21:9

Video Generation

POST /video-generation
Content-Type: application/json

{
  "prompt": "Product rotating on display stand",
  "image_url": "https://example.com/product.jpg",
  "model": "ltx",
  "duration": 5,
  "aspect_ratio": "16:9"
}

Cost: 10-80 credits

Models: veo31, ltx, kling, sora

3D Model Generation

POST /image-to-3d
Content-Type: application/json

{
  "image_url": "https://example.com/product.jpg",
  "model": "meshy",
  "topology": "quad",
  "enable_pbr": true
}

Returns: Multiple formats (glb, obj, fbx, usdz) Cost: 2-50 credits

Alt Text Generation

POST /alt-text
Content-Type: application/json

{
  "image_url": "https://example.com/image.jpg"
}

Cost: 1 credit

File Upload

POST /upload
Content-Type: multipart/form-data

file: [binary data, max 10MB]

Cost: FREE

Batch Operations

POST /batch/remove-bg
POST /batch/upscale
POST /batch/alt-text
Content-Type: application/json

{
  "images": [
    {"id": "1", "image_url": "https://example.com/img1.jpg"},
    {"id": "2", "image_url": "https://example.com/img2.jpg"}
  ],
  "model": "birefnet"
}

Check batch status:

GET /batch/status/{jobId}

Rate Limits

| Tier | Requests/Hour | |------|---------------| | Free | 20 | | Pro | 500 | | Business | 2,000 |

Code Examples

JavaScript

const SIRV_API = 'https://www.sirv.studio/api/zapier';
const API_KEY = 'sk_live_YOUR_API_KEY';

async function removeBackground(imageUrl) {
  const response = await fetch(`${SIRV_API}/remove-bg`, {
    method: 'POST',
    headers: {
      'Authorization': `Bearer ${API_KEY}`,
      'Content-Type': 'application/json'
    },
    body: JSON.stringify({
      image_url: imageUrl,
      model: 'birefnet'
    })
  });
  return (await response.json()).image_url;
}

async function generateImage(prompt, aspectRatio = '1:1') {
  const response = await fetch(`${SIRV_API}/generate`, {
    method: 'POST',
    headers: {
      'Authorization': `Bearer ${API_KEY}`,
      'Content-Type': 'application/json'
    },
    body: JSON.stringify({
      prompt,
      model: 'flux-kontext',
      aspect_ratio: aspectRatio
    })
  });
  return (await response.json()).image_url;
}

cURL

# Remove background
curl -X POST https://www.sirv.studio/api/zapier/remove-bg \
  -H "Authorization: Bearer sk_live_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"image_url": "https://example.com/product.jpg", "model": "birefnet"}'

# Generate image
curl -X POST https://www.sirv.studio/api/zapier/generate \
  -H "Authorization: Bearer sk_live_YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"prompt": "Product photo of headphones", "model": "flux-kontext", "aspect_ratio": "1:1"}'

# Check credits
curl https://www.sirv.studio/api/zapier/me \
  -H "Authorization: Bearer sk_live_YOUR_API_KEY"