AI updates from the past week: OpenAI Codex adds internet access, Mistral releases coding assistant, and more — June 6, 2025
AI Updates from the Past Week: OpenAI Codex Adds Internet Access, Mistral Releases Coding Assistant, and More — June 6, 2025
As the world of artificial intelligence continues to evolve at a breakneck pace, two significant updates have captured the attention of developers and tech enthusiasts alike. OpenAI has enhanced its powerful AI coding agent, Codex, by granting it internet access during task execution, while Mistral has unveiled a robust coding assistant, further expanding the capabilities of AI in software development. These advancements not only highlight the rapid progress in AI technology but also underscore the growing reliance on AI tools in the coding community.
OpenAI Codex: Enhanced with Internet Access
OpenAI's Codex, now available to ChatGPT Plus users for just $20 a month, has been updated to include internet access during task execution[1][4]. This feature allows Codex to install dependencies, run tests that require external resources, and upgrade or install packages needed for coding tasks[2][5]. Internet access is opt-in and configurable, enabling users to control which domains and HTTP methods Codex can access[3]. This update is particularly significant because it opens up new possibilities for dynamic coding tasks, such as fetching data from APIs or pulling code updates from repositories like GitHub.
Key Features of Codex Update:
- Internet Access: Enables tasks like dependency installation and external resource testing.
- Configurable Access: Users can control which domains and HTTP methods Codex uses.
- Availability: Available to Plus, Pro, and Team users, with Enterprise access coming soon[3][5].
As someone who's followed AI for years, it's fascinating to see how tools like Codex are bridging the gap between human developers and AI assistants. This integration not only streamlines coding processes but also raises questions about the future of software development—will we see a shift towards more AI-driven coding environments?
Mistral Code: A New Player in AI Coding Assistants
Mistral has entered the AI coding assistant market with Mistral Code, building on the open-source project Continue. This platform is designed to provide enterprise developers with a comprehensive suite of coding models, including Codestral, Codestral Embed, Devstral, and Mistral Medium[2]. Mistral Code is proficient in over 80 programming languages and can reason over files, Git diffs, terminal output, and issues. Currently available as a private beta for JetBrains IDEs and VSCode, Mistral Code aims to deliver instant completions and multi-step refactoring through an integrated platform[2].
Key Features of Mistral Code:
- Multi-Language Support: Proficient in over 80 programming languages.
- Comprehensive Analysis: Can reason over various coding elements like files and Git diffs.
- Private Beta: Available for JetBrains IDEs and VSCode.
Mistral's approach is noteworthy because it emphasizes an integrated platform where all parts of the stack, from models to code, are delivered by a single provider. This unified approach could simplify the deployment and management of AI coding tools within enterprises.
Future Implications and Potential Outcomes
The integration of internet access into tools like Codex and the emergence of new AI coding assistants like Mistral Code signal a significant shift in how AI is used in software development. These advancements could lead to more efficient coding processes, potentially reducing the time and effort required for development tasks. However, they also raise important questions about data security and the ethical use of AI in coding environments.
By the way, the stateless nature of Codex, which requires users to provide clear and complete prompts for each session, highlights the importance of user input in AI-driven coding[4]. This aspect underscores the collaborative role AI plays in development rather than a replacement for human developers.
Comparison of AI Coding Tools
| Feature | OpenAI Codex | Mistral Code |
|---|---|---|
| Internet Access | Yes, opt-in and configurable | Not specified |
| Language Support | Not specified | Over 80 languages |
| Availability | ChatGPT Plus, Pro, Team users | Private beta for JetBrains IDEs and VSCode |
| Key Functionality | Dependency installation, external resource testing | Instant completions, multi-step refactoring |
| Integration | GitHub connection available | Integrated platform for enterprise use |
Interestingly enough, both OpenAI and Mistral are pushing the boundaries of what AI can achieve in coding. While Codex focuses on dynamic task execution with internet access, Mistral Code offers a comprehensive suite of coding models for enterprise developers.
Conclusion
In conclusion, the past week has seen significant advancements in AI coding tools, with OpenAI's Codex gaining internet access and Mistral unveiling its robust coding assistant. These updates highlight the rapid evolution of AI in software development, promising more efficient and dynamic coding environments. As these technologies continue to develop, it will be crucial to address the ethical and security implications of integrating AI into coding workflows.
Excerpt: OpenAI's Codex now includes internet access, while Mistral introduces a new coding assistant, marking significant advancements in AI-driven software development.
Tags: artificial-intelligence, machine-learning, openai, mistral-code, ai-coding-assistants
Category: Core Tech - artificial-intelligence