Resemble AI Documentation

Resemble AI provides a suite of tools to build authenticity, provenance, and identity workflows across audio, images, video, and documents:

• Deepfake Detection: Analyze media for synthetic or manipulated content.
• Detect Agents: Run managed investigations and retrieve history for specific agents.
• Intelligence: Extract claims, entities, context, and risk signals from media.
• Detect Intelligence: Use natural-language queries to analyze completed detection reports.
• Audio Source Tracing: Identify the likely origin of synthetic audio.
• Identity API: Enroll people/brands, attach reference media, and search for matches.
• Watermarking: Apply and detect resilient watermarks across audio, images, and video.
• Secure Uploads: Send private media to analysis jobs without public hosting.

Resemble AI offers APIs for voice generation, understanding, and management:

• Generation & Transformation: Text-to-Speech, Speech-to-Speech, and Streaming over WebSocket.
• Understanding & Enhancement: Speech-to-Text, Audio Enhancement, and Transcript Intelligence.
• Creation & Management: Voice Cloning, Recordings, and Voice Design.
• Asset Management: Use the Clips API to manage generated output and the Projects API to organize related work.

Developers should refer to the following core resources to begin building:

• Developer Quick Start: Guidance on configuring the SDK and making the first request.
• Authentication: Instructions on securing requests using API keys.
• Rate Limits: Information on throughput, retries, and scaling behavior.
• Error Handling: Guidance on interpreting API failures and implementing recovery paths.