Elastic Compute Service (ECS)
Lingma
Unlock the Power of AI for Your Business
Free Trial
1 Million Free Qwen Tokens
Cost-effective
88% Price Reduction Now
Qwen3: Major Models Unveiled! Compact yet powerful—runs smoothly on everyday devices!
Qwen3: Major Models Unveiled!
All Qwen3 models seamlessly integrate "thinking" and "non-thinking" modes, allowing you to switch modes during conversation.
Try NowInnovative Dual Thinking Modes
Balancing inference depth and speed flexibly while managing costs effectively with Qwen3's unique "Thinking" (deep reasoning) and "Non-thinking" (fast response) modes
Lightweight Powerhouse
Excelling in versatility and task execution accuracy among same-sized instruct models
Agent Enhancement
Significant enhancement of Agent capabilities with accurate tool use for advanced automation workflows
Multilingual Excellence
Supporting over 100 languages and dialects for highly reliable comprehension, translation, and reasoning
Apply AI to Your Business Now
Start transforming your business today
Contact ExpertDeepSeek Deployment
Intelligent Design
AI Agents
Multimodal Recognition
One-click, low-cost, rapid deployment of DeepSeek model
Users can deploy the DeepSeek model with minimal effort. It supports various model sizes, automatically scales to meet demand, and is perfect for businesses or projects needing flexible, cost-effective AI solutions.
Technical Implementation Reference
Log on to the PAI console. Select a region in the upper-left corner. In the left-side navigation pane, choose Workspaces and click the name of the desired workspace. In the left-side navigation pane, choose QuickStart > Model Gallery.
On the Model Gallery page, find the model you want to deploy such as DeepSeek-R1-Distill-Qwen-32B. Click the model card.
Click Deploy in the upper-right corner, select the deployment method and resources. Then, click Deploy to create a service in Elastic Algorithm Service (EAS) of PAI.
Related Products
Choose the Right Qwen Model for You at the Best Price
Only the pricing of Model Studio's flagship models is displayed. For more information on pricing and strategies for other models, please check model list.
Model ListFlagship Models on Model Studio
Introduction
Qwen-Max, Qwen-Plus-Latest, and Qwen-Turbo-Latest are flagship models on Model Studio, which are suitable for various scenarios such as intelligent customer service, text creation (such as manuscript writing and copywriting), text editing, and summarization. The scenarios for each model are described in detail below. If you're still not sure which model to choose, it's recommended to try Qwen-Plus-Latest first, which is relatively balanced in terms of performance, speed, and cost.
Type
Qwen-Turbo-Latest
Qwen-Plus-Latest
Qwen-Max
Scenarios
Fast, cost-effective LLM (dynamic update)
Suitable for scenarios requiring high response speed and cost-effectiveness (e.g., customer service conversations, simple Q&A, text generation).
Suitable for scenarios requiring high response speed and cost-effectiveness (e.g., customer service conversations, simple Q&A, text generation).
Balanced LLM (dynamic update)
Suitable for scenarios requiring higher accuracy and logical reasoning (e.g., code generation, data analysis, complex Q&A).
Suitable for scenarios requiring higher accuracy and logical reasoning (e.g., code generation, data analysis, complex Q&A).
Most Capable Billion-Scale LLM
Provides the best inference performance among Qwen models, especially for complex and multi-step tasks.
Provides the best inference performance among Qwen models, especially for complex and multi-step tasks.
Reasoning Skills
No.3
No.2
No.1
Response Speed
No.1
No.2
No.3
Price per Thousand Tokens
Note
All three models support API access. For more information, see Qwen API reference 
AI Transforming SMB Industries
Leverage AI to drive growth and efficiency in your business
Contact AI ExpertFinance
Secure and Cost-Effective AI Solutions for Financial Businesses
Safeguard your data while accelerating intelligent transformation with Alibaba Cloud’s tailored AI solutions. Financial SMBs often face the dual challenge of maintaining strict data security and system reliability, while also managing the costs of AI-driven upgrades. Our solution enables you to deploy custom AI models on high-performance servers, ensuring secure, private networking that seamlessly integrates with your internal systems. This approach not only meets compliance requirements but also delivers efficient data processing, empowering your business to enhance decision-making, reduce risks, and stay competitive. Powered by Alibaba Cloud’s advanced infrastructure, you can achieve a secure, cost-effective, and scalable AI transformation that drives growth without compromising safety.
Contact AI ExpertGaming
Level Up Your Gaming Business with AI-Powered Tools on Alibaba Cloud
Looking to build better games faster and keep your players happy? Alibaba Cloud makes it easy for small and medium gaming businesses like yours to speed up development and improve operations with AI-powered tools. Imagine having a code assistant that helps your team write cleaner, error-free code faster—or an AI customer service assistant that resolves player queries in real-time. Plus, our AI-driven operation tools can monitor your game’s performance and catch issues before they impact players. All of this is powered by Alibaba Cloud’s Deepseek R1 on the EGS platform , giving you the speed and scalability you need to focus on what matters most. Ready to take your game to the next level?
Contact AI ExpertAI & Cloud Computing
Effortlessly Handle High AI Workloads
For AI-native businesses, managing sudden spikes in user traffic can be daunting—requiring millisecond-level AI responses, stable high-load performance, and efficient resource allocation across multiple models. Alibaba Cloud’s Model Studio rises to the challenge, supporting over 5 million TPM (transactions per minute) to meet high-concurrency demands. It ensures fast, reliable access to the Qwen series models , while safeguarding data security and compliance. With this powerful technology, you can deliver seamless global services, optimize resource usage, and keep your business running smoothly at scale.
Contact AI ExpertEdTech
Go Global with AI-Powered Multilingual Education Solutions
Break language barriers and expand your online education platform globally with Alibaba Cloud. Our advanced Qwen-VL+Turbo model simplifies the complexities of multilingual content, including right-to-left layouts and diverse character recognition, while handling image-text parsing with ease. Powered by the Model Studio platform , you can quickly develop AI-driven education tools that support multiple languages, ensuring seamless learning experiences for students worldwide. Launch faster, reduce technical challenges, and unlock new markets—all while focusing on what matters most: delivering quality education. Let AI take your platform global—effortlessly.
Contact AI ExpertAI Ready Cloud Products
Products
Model Studio
A one-stop generative AI platform for building intelligent applications that understand your business, utilizing the Qwen model series such as the MoE model Qwen-Max.
Enhanced foundation models
Built-in model inference and evaluation pipelines
Simplified GenAI application development
Comprehensive security measures
Enhanced foundation modelsBuilt-in model inference and evaluation pipelinesSimplified GenAI application developmentComprehensive security measuresProducts
Platform for AI
A platform that provides enterprise-grade data modeling services powered by machine learning algorithms, designed to kickstart your data-driven operations.
Deploy DeepSeek-V3 and DeepSeek-R1 models
Deploy an LLM as an application in PAI-EAS
Deploy Stable Diffusion in PAI-EAS with a few clicks
Build an AI chatbot in PAI-EAS
Use EasyVision to detect objects
Deploy DeepSeek-V3 and DeepSeek-R1 modelsDeploy an LLM as an application in PAI-EASDeploy Stable Diffusion in PAI-EAS with a few clicksBuild an AI chatbot in PAI-EASUse EasyVision to detect objectsProducts
Elastic GPU Service
Elastic computing instances with GPU accelerators, suitable for applications in artificial intelligence (including deep learning and machine learning), high-performance computing, and professional graphics.
Industry-leading solution with high performance
High-quality video processing
Ideal for high-performance computing
Industry-leading solution with high performanceHigh-quality video processingIdeal for high-performance computingProducts
Elasticsearch
EIPs are independent public IPs, decoupling them from ECS instances for easier management.
New AI search solution with Elasticsearch
AI search integrated with LLM
High-relevance semantic search
Hybrid information retrieval
New AI search solution with ElasticsearchAI search integrated with LLMHigh-relevance semantic searchHybrid information retrievalProducts
AnalyticDB
AnalyticDB is a cloud-native real-time data warehouse that features storage-compute decoupling. It integrates seamlessly with MySQL, PostgreSQL, GenAI, and open data lakehouse ecosystems.
Compatibility with various data ecosystems
One-stop high-performance real-time analysis
Seamless AI integration
High elasticity & cost-effectiveness
Compatibility with various data ecosystemsOne-stop high-performance real-time analysisSeamless AI integrationHigh elasticity & cost-effectivenessProducts
Object Storage Service (OSS)
Fully managed object storage service to store and access any amount of data from anywhere
Industry-leading SLA
Cost-effective and easy-to-consume
Enterprise-ready security and compliance
Easy data migration
Industry-leading SLACost-effective and easy-to-consumeEnterprise-ready security and complianceEasy data migrationGet AI Training And Upskill with Our Generative AI Certification!
Boost your AI skills with Alibaba Cloud's certification program. Perfect for AI enthusiasts, aspiring professionals, developers, data scientists, and visionary tech leaders. Enroll now to stay ahead in the AI revolution.
