Multilingual voice agents
Build assistants that can listen, speak, see an image or camera frame and call application tools within the same experience.
Independent tool overview
Qwen3.5-Omni at a glance
Qwen3.5-Omni is Alibaba's hosted multimodal model family for applications that need to understand text, images, audio and video within one conversation and respond with text or synthesized speech.
Visit the official Qwen3.5-Omni site ↗
- Inputs
- Text, images, audio and video
- Outputs
- Text and synthesized speech
- Context window
- 262K tokens on QwenCloud
- Speech input
- 113 languages and dialects
- Speech output
- 36 languages and dialects
- Access
- Hosted API through Alibaba Cloud Model Studio or QwenCloud
Overview
What Qwen3.5-Omni is
Qwen3.5-Omni combines visual, audio, video and language understanding in one hosted model family. The standard Plus and Flash APIs suit asynchronous analysis and speech generation, while separate Realtime variants are designed for low-latency voice and camera experiences.
Its broad modality support makes it useful for multilingual assistants, media analysis and accessibility workflows, but teams should evaluate the exact endpoint they plan to ship. Plus, Flash and Realtime differ in price, latency and supported tools.
Qwen3.5-Omni should not be confused with the older Qwen3-Omni open-weight release. Qwen3.5-Omni is currently documented as a hosted Model Studio and QwenCloud API; do not assume the older model's Apache 2.0 license applies.
Use cases
Who Qwen3.5-Omni is best for
The strongest fit depends on the job you need the product to complete, not the size of its feature list.
Audio and video analysis
Create transcripts, scene descriptions, speaker-aware summaries and searchable metadata from long-form media.
Multimodal product workflows
Combine screen, camera, voice and text context for support, training, accessibility or creative applications.
Capabilities
Core Qwen3.5-Omni features
Native multimodal understanding
Processes text, images, audio and video without requiring a separate model for every input type.
Text and speech responses
Can return text alone or stream text with generated audio using supported voices and endpoints.
Long-media support
Official materials describe more than 10 hours of audio and more than 400 seconds of 720p video at one frame per second for supported workflows.
Realtime variants
Dedicated Plus and Flash Realtime models support low-latency speech-to-speech and audio-visual interactions.
Tools and web search
Supported HTTP and WebSocket configurations can use web search, while function-calling availability depends on the endpoint.
Fine-grained voice control
Supports multilingual speech output and expressive controls; voice-cloning availability and consent requirements should be checked before use.
Process
How the Qwen3.5-Omni workflow works
Step 1
Choose the exact endpoint
Select Plus, Flash or Realtime based on whether accuracy, price or conversational latency matters most.
Step 2
Constrain media and permissions
Set input limits, retention rules and consent requirements before sending private recordings, screens or voices.
Step 3
Evaluate representative inputs
Test the languages, accents, noise levels, camera conditions and media lengths your users will actually provide.
Step 4
Add operational safeguards
Monitor token growth, tool calls, latency and failure cases, and provide a human or text fallback for critical workflows.
Cost
Qwen3.5-Omni pricing and free plan
Qwen3.5-Omni is metered by modality. Current international QwenCloud rates differ for text/image/video, audio and speech output, and multi-turn realtime sessions can rebill retained conversation context. Alibaba Cloud region and deployment pricing may vary.
Qwen3.5-Omni Plus
$1.40–$44 per 1M tokens
Higher-capability hosted model with modality-specific rates.
- $1.40 per 1M text, image or video input tokens
- $11 per 1M audio input tokens
- $8.30 per 1M text output tokens
- $44 per 1M audio output tokens; accompanying output text is not charged
Qwen3.5-Omni Flash
$0.40–$11.90 per 1M tokens
Lower-cost hosted variant with the same broad modality categories.
- $0.40 per 1M text, image or video input tokens
- $3 per 1M audio input tokens
- $2.20 per 1M text output tokens
- $11.90 per 1M audio output tokens
Provisioned Flash deployment
From $7.014 per hour
Dedicated Alibaba Cloud deployment option for stable throughput.
- Listed at $3,383.024 per month for one MU9 unit
- Deployment region and contract terms can change availability
Pricing checked . Check current pricing at the source ↗
Assessment
Qwen3.5-Omni strengths and limitations
Where it stands out
- One model family covers text, visual, audio and video inputs plus speech output
- Wide documented speech-language coverage for global applications
- Long audio and video understanding can simplify media pipelines
- Flash, Plus and Realtime options let teams trade capability against cost and latency
What to consider
- Qwen3.5-Omni is a hosted offering and should not be presented as the older open-weight Qwen3-Omni model
- Plus does not support a deep-thinking mode, and tool support differs between HTTP and realtime endpoints
- Video is sampled and endpoint-specific frame limits apply, so the nominal context window is not equivalent to lossless long-video recall
- Realtime conversations can become progressively more expensive because retained history is processed again on later turns
- Support for 113 languages and dialects does not guarantee equal accuracy across accents, domains or noisy audio
Compare
Qwen3.5-Omni alternatives
The right alternative depends on the specific output, workflow, controls and budget your project requires.
Business Operations
Gemini 3.1 Flash Live
Consider it for Google's low-latency native audio and video agent stack.
Explore Gemini 3.1 Flash Live →Miscellaneous
GPT-Realtime-2
Consider it for OpenAI-based realtime voice agents, tool use and interruption handling.
Explore GPT-Realtime-2 →Miscellaneous
Voxtral Transcribe 2
Consider it when accurate live or batch transcription matters more than full multimodal generation.
Explore Voxtral Transcribe 2 →Questions
Qwen3.5-Omni FAQs
Is Qwen3.5-Omni open source?
Qwen3.5-Omni is currently documented as a hosted API family. The older Qwen3-Omni release has open weights under Apache 2.0, but that license should not be assumed to cover Qwen3.5-Omni.
What can Qwen3.5-Omni process?
Supported endpoints accept text, images, audio and video and can return text or synthesized speech. Realtime variants are available for interactive voice and camera experiences.
How much does Qwen3.5-Omni cost?
International QwenCloud pricing starts at $1.40 per million text, image or video input tokens for Plus and $0.40 for Flash. Audio input and audio output cost more, and regional Alibaba Cloud rates can differ.
Does Qwen3.5-Omni support function calling?
Supported HTTP configurations can call functions. Some realtime WebSocket configurations do not, so verify the capability table for the exact model ID before designing an agent workflow.
Can it understand long videos?
Official materials describe more than 400 seconds of 720p video at one frame per second for supported workflows. Sampling and endpoint frame limits mean teams should test scene-level recall on their own content.
Bottom line
Our Qwen3.5-Omni verdict
Qwen3.5-Omni is a strong hosted option when one application needs multilingual speech, visual context and media understanding together. Its practical value depends on selecting the right Plus, Flash or Realtime endpoint and controlling modality-specific costs, privacy and long-context behavior.
Visit Qwen3.5-Omni website ↗Similar Consumer Tools
View all Consumer AI tools →
Cohere Transcribe
Cohere Transcribe- SOTA, free open-source speech recognition model
Trinity-Large-Thinking
Trinity-Large-Thinking - Arcee AI's new open-weight frontier reasoning model for long-horizon agents
Minimax M2.7
MiniMax M2.7 - MiniMax's new 'self-evolving' model with strong coding and agentic benchmarks
LFM2.5-350M
LFM2.5-350M - Liquid AI's 350M-parameter edge model built for tool use and on-device agents
AI training for the future of work.
Get access to all our AI courses, hundreds of real-world AI use cases, live expert-led workshops, an exclusive network of AI early adopters, and more.
AI Courses
Get unlimited access to all of our current & upcoming industry-specific AI courses for the duration of your subscription.
Daily Guides
To keep up with the rapid pace of AI, our team publishes AI implementation guides daily. Our library contains 300+ practical use cases to automate real-world work.
Workshops
Join weekly, live, interactive sessions with industry leaders who are at the forefront of AI for hands-on implementation guidance and exclusive insights.
Community
Network with an exclusive community of AI-first professionals who are working smarter with AI. Learn how early adopters are using AI in their work and businesses.