Table of Content
OpenAI's late night show! GPT-4.1 series comes with mini/nano, API is free for a limited time!
Updated on:July-01st-2025
Recommendation
OpenAI leads the new era of AI, and the GPT-4.1 series of models are launched with limited-time free API!
Core content:
1. The new GPT-4.1 series of models are released, and they are fully available through API calls
2. Performance improvement: coding ability, long text understanding, etc. are significantly enhanced, and the response speed is faster
3. Limited-time free benefits: From April 15 to 22, 2025, the API calls of the three new models are completely free
Yang Fangxian
Founder of 53AI/Most Valuable Expert of Tencent Cloud (TVP)
1. New model release:gpt-4.1 series
- Model List:
gpt-4.1gpt-4.1-minigpt-4.1-nano- How to obtain:
The calls are fully available through API. - Core advantages :
It has improved performance in coding ability, long text comprehension, etc., with faster response speed and better cost of use. - Limited time free benefits:
- Free time:
April 15, 2025 - April 22, 2025 - Scope of application:
During this period, the three new models are completely free to use via the API . - Free tier available:
Free tier users can also enjoy a free trial with the following rate limits: - 3 requests per minute (RPM = 3)
- 200 requests per day (RPD = 200)
2. Old model deprecation notice
- Deprecated models:
gpt-4.5-preview - Current status:
This model has been officially marked as Deprecated . - Removal Date:
Will be removed from the API on July 14, 2025 .
Related PostsView More
DeepSeek R1 0528 made me rethink AI programming
Explore DeepSeek R1 0528's redefinition of AI programming! This model allows everyone to easily reach the stars of AI programming. Its one-time generated front-end cases such as Tetris and commodity inventory management are rich and colorful. During use, we found that the prompt writing for DeepSeek R1 0528 is different from the non-inference model, and it is necessary to find the right prompt...
OpenMemory MCP: Giving AI Tools "Shared Memory"
Explore how OpenMemory MCP enables AI tools to have "shared memory" and solve the "memory fragmentation" problem caused by switching between different editors. It is a private memory server running on a computer, using a local priority strategy to ensure data privacy. Uncover its working principle, extracting and updating memory through a two-stage intelligent processing flow, as smart as the...
Analysis of the implementation of RAG technology in the RAGFlow project
In-depth discussion of the implementation details of RAG technology in the RAGFlow project! This article summarizes the best practices in deep document understanding, document segmentation, and Embeddings model selection and configuration by analyzing its source code and official documents. As the key to retrieval enhancement generation, RAG technology has demonstrated its strength in RAGFlow....