Tongyi Qianwen

I. Core Functions and Model Matrix

1. Qwen Large Language Model Series

• Qwen3-Max: All-powerful flagship model with the strongest capabilities in natural language understanding, logical reasoning and complex task processing
• Qwen-Plus: Flagship balanced model, balancing performance and cost, suitable for most general scenarios
• Qwen-Flash/Qwen3-Omni-Flash: Lightweight and ultra-fast model with extremely fast response speed, supporting high-concurrency calls
• Qwen3-Coder-Plus: Professional code and AI Agent model, proficient in code writing, debugging and automated task execution
• Qwen3-VL-Plus/Qwen3-VL-Flash: Visual perception model, supporting image understanding, object recognition and image-text interaction
• Qwen-Image: Precision drawing model that can generate high-quality image content in line with instructions
• Specialized Capability Models: Qwen-Doc (long document processing), Qwen-Long (ultra-long text), Qwen-TTS (text-to-speech), CosyVoice (anthropomorphic speech), Fun-ASR (speech recognition)

2. Wanxiang Vision Generation Large Model Series

• Wan2.6-T2I: Text-to-image with strong instruction-following capabilities
• Wan2.6-T2V: Text-to-video, enabling natural audio-visual synchronization
• Wan2.6-I2V: Image-to-video, supporting intelligent multi-shot narrative
• Wan2.6-R2V: Video character reference generation, maintaining character consistency
• Wan2.6-Image: Supporting mixed image-text output
• Wan2.2-Animate: Video face-swapping and image-to-motion generation

3. General Core Capabilities

• Natural Language Understanding & Generation: Conversation, writing, translation, summarization, Q&A
• Multimodal Interaction: Image/video/audio understanding and generation
• Tool Calling & AI Agent: Automated execution of complex tasks
• Data Processing: Entity recognition, information extraction, long document summarization, text tagging
• Security Capabilities: Content moderation, device risk control, internet anti-fraud

II. Free Policy & Paid Plans

1. 1. Personal Free Quota: Basic web-based features are permanently free; lightweight models such as Qwen-Flash and Qwen3-Omni-Flash provide daily free call limits; Qwen-Plus, Qwen3-VL-
   2. Companionship & Social: Virtual IP creation, real-time emotional conversation, intelligent customer service
   3. Smart Cockpit: Travel assistant, route planning, entertainment recommendation, vehicle control
   4. Data Services: Bid document processing, HR resume screening, e-commerce information extraction
   5. Security Risk Control: Content security moderation, device risk control, internet anti-fraud

VI. Core Advantages Over Peers

1. 1. 1. Full-Modal Full-Gradient Model Matrix: Covers all model levels from lightweight ultra-fast to all-powerful ultimate, with full-modal capabilities including large language, vision, audio and code, eliminating the need to integrate models from multiple vendors
      2. Leading Multimodal Generation: The Wanxiang vision large model achieves globally top-tier performance in video generation motion range, physical law compliance and audio-visual synchronization, supporting advanced features such as multi-shot narrative and character reference generation
      3. Outstanding Enterprise-Grade Services: Serving 300,000+ enterprise customers with mature private deployment, customized fine-tuning and technical support systems to meet the needs of enterprises of all sizes
      4. Superior Chinese Language Capabilities: Deeply optimized for Chinese contexts, outperforming overseas large models in Chinese understanding, generation and cultural adaptation
      5. Comprehensive Ecosystem: Deeply integrated with Alibaba Cloud’s cloud computing, big data, security and other products, providing one-stop AI solutions

VII. Usage Notes

1. 1. 1. Content Compliance Requirements: It is prohibited to use Tongyi Qianwen to generate illegal, non-compliant, pornographic, violent or sensitive content; otherwise, the account may be restricted
      2. Data Security Protection: Do not input personal sensitive information (e.g., ID number, bank card number) or enterprise trade secrets in conversations; enterprise users are recommended to use private deployment solutions
      3. Awareness of Model Limitations: Content generated by Tongyi Qianwen may contain factual errors or logical flaws; important information should be verified manually
      4. API Call Specifications: Comply with Alibaba Cloud API call frequency limits to avoid account suspension due to malicious calls
      5. Copyright Issues: When using content generated by Tongyi Qianwen for commercial purposes, comply with Alibaba Cloud’s copyright agreement to ensure no infringement of third-party intellectual property rights

VIII. Frequently Asked Questions (FAQ)

1. 1. 1. Q: What if the free quota is exhausted?

         A: Personal users can wait for the daily free quota reset or upgrade to a paid user for more calls; developers can purchase additional API call quotas on Alibaba Cloud Model Studio.

      2. Q: What languages does Tongyi Qianwen support?

         A: Tongyi Qianwen supports multiple major languages including Chinese, English, Japanese and Korean, with its Chinese capabilities being the most outstanding.

      3. Can generated content be used for commercial purposes?

         A: Content generated by Tongyi Qianwen may be used for commercial purposes subject to Alibaba Cloud’s service agreement and relevant laws and regulations, but users shall bear their own risks regarding content compliance and copyright.

      4. Can the Tongyi Qianwen model be fine-tuned?

         A: Enterprise users can upload their own data on Alibaba Cloud Model Studio for customized fine-tuning of the Tongyi Qianwen model to enhance performance in specific scenarios.

      5. How to handle API call failures?

         A: First, check whether the API key is correct and the call format complies with specifications; if the problem persists, view error logs on Alibaba Cloud Model Studio or contact Alibaba Cloud technical support.