
Tongyi Qianwen
Features of Tongyi Qianwen
Use Cases of Tongyi Qianwen
FAQ about Tongyi Qianwen
QWhich models are available in Tongyi Qianwen?
Official versions include Max, Flash, Omni, Omni-Realtime, and QVQ, optimized for deep reasoning, speed, multimodal, or visual inference respectively.
QHow do I call the API?
Activate Alibaba Cloud Bailian service to get an HTTPS endpoint; Python, Go, and Java SDKs are provided for integration within ten minutes.
QHow is pricing calculated?
Pay-as-you-go by input + output tokens; e.g., Flash costs 0.00015 yuan/1k input tokens and 0.0015 yuan/1k output tokens up to 128k context, with 1 million free tokens for new users.
QDoes it support private deployment?
Yes, dedicated cloud and on-premise appliance deployments keep data local, meeting strict compliance requirements for finance and government sectors.
QWhat is the maximum context length?
Flash supports 1 million tokens and Max supports 250k tokens; enterprises can choose according to business needs.
QHow is content safety ensured?
Built-in Alibaba Cloud Green Network moderation, TLS 1.3 encrypted transmission, and support for Class-III cybersecurity protection and national cryptography standards.
QCan I fine-tune the model?
The console allows lightweight fine-tuning with private corpora and integration with vector databases for RAG to improve domain-specific accuracy.