Alibaba Cloud Model Studio offers Qwen and third-party models for text, image, audio, and video.
Text generation
Complex reasoning and coding
Balanced performance, speed, and cost
Fast and cost-effective
Text generation
Text generation
Text generation
Text generation
Text generation
Text generation
Image & video
Understanding
Extract text descriptions or structured data from images and videos
Visual understanding
Visual understanding
Visual understanding
Visual understanding
Generation
Generate images and videos from text or images, with support for editing, reference, and high-resolution output
Image generation
Image generation
Video generation
Video generation
Video generation
Video editing
Video generation
World models
Real-time interactive open-world model — Adventure, Directing, and Acting
World exploration
Real-time directing
Character acting
Audio & speech
Audio generation
Generate audio content including voices, sound effects, and ambient sounds
Text-to-speech
For audiobook reading, voice broadcasting, virtual avatars, and more
Real-time and non-real-time
Real-time and non-real-time
Speech recognition
Convert speech to text for real-time voice input, continuous transcription, and transcription of recorded audio
Real-time speech recognition
Real-time speech recognition
Non-real-time speech recognition
Non-real-time speech recognition
Speech-to-speech
End-to-end voice conversation without separate ASR and TTS calls
Speech-to-speech
Speech-to-speech
Speech-to-speech
Speech-to-speech
Omni
Integrates understanding and generation across text, images, audio, and video. For audio/video analysis and text generation, use qwen3.8-omni-flash through Chat Completions or Responses. For real-time audio/video conversations, use qwen3.8-omni-flash-realtime.
NoteFor non-real-time audio output, use qwen3.5-omni-plus.
Embeddings & reranking
Convert text or multimodal content into vectors, combined with reranking to improve retrieval accuracy
Text embeddings
Text embeddings
Multimodal embeddings
Reranking
Decision Model
A structured decision model for high-frequency business decisions. It performs classification, yes-or-no decisions, and scoring in a single forward pass, returning probability distributions and confidence.
View all models
Go to Model Plaza to browse all Qwen, third-party, domain-specific, and legacy models.