Protect your media from unauthorized AI training.

Sidearm is an API platform for watermarking, adversarial embeddings, AI content detection, content credentials, neural fingerprinting, and similarity search for images, video, audio, and text. Run SOTA proprietary, and open-source protection algorithms through a single API.

Add Sidearm tools to Cursor, Claude Desktop, Windsurf, or any MCP-compatible agent:

{
  "mcpServers": {
    "sdrm": {
      "command": "npx",
      "args": ["-y", "@sidearmdrm/mcp"],
      "env": { "SDRM_API_KEY": "sk_live_..." }
    }
  }
}

npm install @sidearmdrm/sdk-node

import { Sidearm } from "@sidearmdrm/sdk-node";

const client = new Sidearm({ apiKey: "sk_live_..." });
const job = await client.protect({
  media_url: "https://example.com/artwork.png",
  level: "maximum",
});
const result = await job.wait();

curl -X POST https://api.sdrm.io/api/v1/protect \
  -H "Authorization: Bearer sk_live_..." \
  -H "Content-Type: application/json" \
  -d '{"media_url": "https://example.com/artwork.png", "level": "maximum"}'

• Watermarking -- Invisible ownership marks that survive compression, cropping, reformatting, and print-to-digital round-trips
• Adversarial Embeddings -- Perturbations that poison AI training data and prevent style mimicry
• Content Credentials -- Cryptographic provenance signing with tamper-evident chain of custody
• AI Detection -- Determine whether content was AI-generated across image, video, audio, and text
• Neural Fingerprinting -- Deep embedding-based indexing for semantic and perceptual matching
• Similarity Search -- Find stolen or derivative content across your indexed library at multiple tiers
• Membership Inference -- Prove whether an AI model was trained on your protected content

• Documentation
• API Reference
• Get an API Key
• Algorithm Catalog