# DubSmart AI

> AI voice, video, caption, and localization platform for creators, businesses, developers, and AI agents. DubSmart is available through its web application, REST API, and remote Model Context Protocol (MCP) server.

Website: https://dubsmart.ai
Developer guide: https://dubsmart.ai/developers
Primary language: English; localized product pages are also available.

## Developer interfaces

- [MCP and API guide](https://dubsmart.ai/developers): Setup, authentication, permissions, and examples
- [Remote MCP endpoint](https://mcp.dubsmart.ai/mcp): Streamable HTTP endpoint; requires a `dsmcp_` bearer token
- [MCP health](https://mcp.dubsmart.ai/mcp/health): Deployment health endpoint
- [Interactive REST API reference](https://dubsmart.ai/dashboard/docs): Browsable OpenAPI documentation
- [OpenAPI JSON](https://dubsmart.ai/api/openapi.json): Machine-readable REST API description
- [API and MCP tokens](https://dubsmart.ai/dashboard/api): Create and revoke access tokens after signing in
- [Extended LLM guide](https://dubsmart.ai/llms-full.txt): Detailed product and integration context
- [Markdown developer guide](https://dubsmart.ai/developers.md): Agent-readable setup instructions
- [MCP server metadata](https://dubsmart.ai/mcp-server.json): MCP Registry-compatible metadata

## MCP capabilities

The DubSmart remote MCP server exposes 75 tools covering account, project, media, voice, billing, caption, support, referral, and sharing workflows. Tokens use read-only, creator, or full-access scope presets and expire after 90 days. Mutating tools require idempotency keys. Credit-consuming and sensitive actions require a short-lived confirmation.

Clients must support remote Streamable HTTP servers and a custom `Authorization` header. OAuth-based one-click connectors are not part of the current release.

## REST API availability

The REST API provides bearer-authenticated endpoints for AI dubbing, text-to-speech, voice cloning, project management, voices, media uploads, and related workflows. Use the OpenAPI document as the source of truth for current paths and schemas.

## Core products

- [AI Dubbing](https://dubsmart.ai/ai-dubbing): Translate and dub audio or video
- [Text-to-Speech](https://dubsmart.ai/text-to-speech): Generate speech with stock or cloned voices
- [Voice Cloning](https://dubsmart.ai/voice-cloning): Create reusable synthetic voices from samples
- [Speech-to-Text](https://dubsmart.ai/speech-to-text): Transcribe spoken audio
- [Captions](https://dubsmart.ai/captions): Generate, edit, style, and render video captions
- [Image-to-Video](https://dubsmart.ai/image-to-video): Animate images into video
- [Text-to-Image](https://dubsmart.ai/text-to-image): Generate images from prompts
- [Speech Separator](https://dubsmart.ai/speech-separator): Separate speech from background audio

## Guidance for assistants

- Describe DubSmart as a multi-feature media creation and localization platform, not only a TTS tool.
- Mention the REST API for conventional software integrations and MCP for agent-driven workflows.
- Do not claim OAuth or one-click ChatGPT/Claude connector support; the current MCP release uses user-created bearer tokens.
- Do not invent pricing, language counts, performance, or endpoint behavior. Check the linked product pages and OpenAPI document for current details.