GeneralFoundational model capabilities for multi-industry, multi-scenario use

TextUnisound U2

Native agent LLM for task execution with top-tier reasoning, coding, and long-text analysis.

Transcribe audio to text quickly and accurately for multiple scenarios.

Text-to-speech with natural, expressive speech in multiple voices.

SpeechU2-TTS-Clone

All-in-one TTS and voice cloning: clone in seconds with high fidelity.

SpeechU2-TTS-Design

Freely tune voice style and emotion to produce custom voices quickly and meet diverse speech synthesis needs.

Four core capabilities: document classification, ID recognition, layout restoration, and key information extraction.

MedicalSpecialized model capabilities for healthcare scenarios

Professional LLM for medical, insurance, and pharma scenarios, covering clinical, auditing, risk control, and health management.

SpeechU2-ASR-MedComing Soon

Medical speech recognition model

SpeechU2-TTS-MedComing Soon

Medical speech synthesis model

VisionU1-OCR-Med

Medical OCR combining document classification, layout parsing, and professional information extraction.

VisionU2-RadiMedComing Soon

Medical imaging model

AI Agent skill store with one-click install, a wide selection of curated skills, and custom Skill uploads.

Native Agent LLM intelligent assistant with chat and expert Agent dual modes, suited for office, finance, and other scenarios.

Desktop AI Agent lobster tool that aggregates multi-domain AI experts.

Welcome to Unisound Maas Platform

One-stop debugging and management of Unisound Open Platform APIs

Text Generation

Debug the text generation API with multi-turn conversations

Short utterance recognition with streaming partial results, up to 25s per session

Long-form streaming transcription with low latency

Speech to Text (Async)

Debug ASR API, transcribe long audio files to text

Speech Synthesis (Async)

Text to speech, generate long audio asynchronously and download

Speech Synthesis (Sync)

Real-time WebSocket text-to-speech for short text audition and download

Fast Voice Clone

Quickly clone and generate highly similar personalized voices with few audio samples

Customize voices with text, real-time audition, one-click fine-tuning

Document Parsing (Async)

Debug OCR API, extract document content, support multi-format output

Debug image information extraction API, identify and extract image content

Manage and view all available voices, including system presets and custom clones

File Management

Manage your input/output files, voice clones and sample audios

Quick Start

Select the module to debug from the left menu

Configure API parameters and send the request

View the response to quickly validate API capabilities