What is Zhipu Qingyan
On August 31, 2023, Zhipu AI officially launched its first generative AI assistant — "Zhipu Qingyan". This assistant is based on Zhipu AI's self-developed bilingual Chinese-English dialogue model ChatGLM-4, pre-trained on trillions of characters of text and code, and uses supervised fine-tuning technology to provide users with intelligent services in the form of general conversation.
As the user'sintelligent dialogue assistant, "Zhipu Qingyan" can answer various questions and complete various tasks for users in work, study, and daily life. At present, Zhipu Qingyan already has rich capabilities such as "general Q&A, multi-turn dialogue, creative writing, code generation, and virtual dialogue", and has opened multimodal generation capabilities: text-to-image, text-to-video, image-to-video, etc.
The newly launched "Research Mode" of Zhipu Qingyan can search and integrate information from the entire web, quickly generate structured and citable research reports, and is suitable for various scenarios such as academic research, business insights, life decisions, and hot topic tracking, helping users analyze problems efficiently and in depth and form clear insights and conclusions.
"Zhipu Qingyan" has been launched in major app stores. Users can download it through the Apple App Store and mainstream Android stores (including Huawei, OPPO, vivo, Xiaomi, etc.), or search for "Zhipu Qingyan" in WeChat Mini Programs to experience its features. On "Zhipu Qingyan", users can enjoy more intelligent, more convenient, and more efficient personalized content acquisition anytime and anywhere.
In addition to developing commercial AI products such as Zhipu Qingyan andQingying, Zhipu AI is also very open in model open source and API, for example:
- CogVideoX(AI video generation model)
- CogView4(AI text-to-image model)
- GLM-4V-Plus(multimodal vision model)
- GLM-4-long(2 million character long-text model)
- GLM-4-Plus(flagship large model)
- GLM-4-Flash(first free language model API)
- GLM-4V-Flash(first free multimodal model API)
- GLM-Zero(AI deep reasoning model)
Main features of Zhipu Qingyan
- Video generation：Zhipu Qingying has AI video generation capabilities and supports two modes: text-to-video and image-to-video.
- Image generation: Zhipu Qingyan supports AI painting functionality, generating exquisite images through simple text descriptions.
- PPT generation: Zhipu Qingyan supports AI PPT generation functionality. Through the "Qingyan PPT" agent, beautiful PPTs can be generated with one click.
- Video call: Zhipu Qingyan is no longer limited to text and voice interaction, but can "see" the world through video, understand users' expressions and emotions, and provide a more natural and more humanized interaction experience.
- Web search: Supports advanced web connectivity, connects to content across the entire web, searches precisely, and is an intelligent assistant for rapid analysis and summarization.
- Multi-turn dialogue: Has excellent dialogue capabilities, can engage in natural and fluent multi-turn dialogue with users, and maintains contextual coherence.
- Creative writing: Provides users withAI writingservices, including article creation, news topic selection, etc., improving writing efficiency and quality.
- Code generation: Supports over 100 programming languages, helps users generate code, answer programming questions, and provide programming suggestions.
- Virtual Conversation: Can play different roles according to user needs, enhancing interactivity and user experience.
- Inspiration Collection: Provides requirement templates for 300+ scenarios, covering multiple vertical fields, helping users quickly obtain writing inspiration.
- Agent Creation: Users can create their own agents, personalized to adapt to specific scenarios or needs, such as teaching assistance, office assistants, etc.
- PPT Mode: One-click generation of professional-grade PPT, supporting personalized styles, color schemes, and layouts, adapting to multiple themes and scenarios.
- Creative Mode: Supports generating creative design content such as posters, long images, web pages, etc., meeting diverse design needs.
- Research Mode: Quickly integrates information from the entire web, generates structured research reports, helping users efficiently and deeply analyze issues, suitable for multiple scenarios such as academic, business, and life decision-making.
How to Use Zhipu Qingyan
- Product Official Website: Visit the Zhipu Qingyan official website, register and log in to your account, or download and install the Zhipu Qingyan APP from the app store.
- Start a Conversation: After logging in, directly enter your question or request in the chat interface, and Zhipu Qingyan will provide an answer or execute the task.
- Use Agents: Explore the agent center, select an agent suitable for your needs, or create a new agent to optimize specific scenarios.
- Upload Documents: If you need AI to interpret documents, upload files in TXT, PDF, Word, and other formats, and ask specific questions or make requests.
- Custom Agents: According to personal or professional needs, customize and create agents, and upload relevant textbooks or materials to enhance their functions.
Suitable Audiences for Zhipu Qingyan
- Students: Can use Zhipu Qingyan to obtain learning materials, answer academic questions, assist with paper writing, etc.
- Teachers: Use Zhipu Qingyan for teaching assistance, course content creation, student assignment grading, etc.
- Researchers: Conduct scientific literature retrieval, data analysis, paper writing, and other research activities.
- Writers and Content Creators: Used for creative writing, copywriting, article editing, etc.
- Programmers and Developers: Obtain programming help, code generation, technical problem solving, etc.
- Corporate Employees: For daily work assistance, document processing, meeting minutes organization, etc.
- Marketing Personnel: Used for marketing copywriting, advertising creativity, social media content generation, etc.
