Refact.ai November Changelog: Qwen-Model Availability, Image Uploads, Self-hosted Version Updates
November is packed with exciting updates. Check out what’s new, and don’t forget to upgrade Refact.ai to the latest version to access them!
- Image Uploads: Add images directly in your IDE with a single click.
- Qwen Model Integration: Get powerful code generation, repair, and reasoning with the new Qwen2.5-Coder series.
- Updates in the Self-hosted Version: Explore new models and integrations.
- Industry Recognition: Refact.ai is proud to be acknowledged among the top AI software solutions for 2024 by Tekpon.
- Next AMA on December 12th: Join us to see how we’ll turn napkin sketches into a website with the Refact.ai Agent.
- Pro Plan Access: Get 1 month of Refact.ai Pro free with the promo code
NOVEMBER.
Let’s dive into the details!
🖼️ Upload Images within Your IDE
Save time with our new in-IDE image upload feature—seamlessly integrated for your convenience. Easily upload one or multiple images to streamline your workflow. How exactly can it help you?
- Automated UI to code: Upload simple UI mockups to automate component creation or small chunks of wireframes directly in your IDE. Let Refact.ai generate frontend code (HTML, CSS, JS) for frameworks like React, Vue.js, or Angular.
- Instant understanding of complex codebases: Upload architecture diagrams, process flows, or system overviews into your workspace to quickly grasp the structure of large or unfamiliar codebases.
- Enhanced visual documentation: Add flowcharts, diagrams, or annotated screenshots directly within your workspace to clarify workflows, highlight key issues, or document processes for future reference.
- Quick visual debugging: Simply upload an image of the bug and let Refact.ai fix it.
It’s already available on Visual Studio Code and JetBrains. Try it first!
🌏 Try Qwen models in Refact.ai
Auto-completion in Refact.ai is becoming increasingly powerful, especially now with the Qwen2.5-Coder series of models. According to their official website:
- Code Generation: The flagship model, Qwen2.5-Coder-32B-Instruct, has achieved the best performance among open-source models on several popular code generation benchmarks (EvalPlus, LiveCodeBench, BigCodeBench) and exhibits competitive performance compared to GPT-4o.
*Source: qwen2.org- Code Repair: Qwen2.5-Coder-32B-Instruct can help users fix errors in their code, making programming more efficient. Aider is a popular benchmark for code repair, and Qwen2.5-Coder-32B-Instruct scored 73.7, performing comparably to GPT-4o on Aider.
- Code Reasoning: Code reasoning refers to the model’s ability to learn the process of code execution and accurately predict the model’s inputs and outputs. The recently released Qwen2.5-Coder-7B-Instruct has already shown impressive performance in code reasoning, and this 32B model takes it a step further.
At Refact.ai, we are committed to staying at the forefront of AI development by integrating the latest models into our product, enhancing its power and capabilities. You can now use the Qwen2.5-Coder family within Refact.ai in the self-hosted version.
And the Refact.ai cloud version now features code completion powered by Qwen2.5-Coder-1.5B. This model stands out as a powerful, low-latency solution, making it an ideal choice for developers seeking a model that combines high performance with efficiency.
🖥️ Other Updates in the Self-hosted Version
- New models are available: Besides the Qwen2.5-Coder family, you can now use the latest Llama3.1 and Llama3.2 within Refact.ai.
- Expanded provider support: Integrate with 3rd party APIs for Groq and Cerebras.
- Configure different models for single-line and multiline code completion: Assign a fast, lightweight model for quick single-line edits, while leveraging a more advanced model for multiline scenarios.
Are you a company that wants to keep your data completely under your control and deploy Refact.ai on your own servers? Fill out the form, and we’ll get in touch with you!
🏆 Refact.ai Recognized as a Top Artificial Intelligence Software for 2024 by Tekpon
Tekpon has announced its list of “Top Artificial Intelligence Software for 2024.” This carefully curated list features tools that showcase the power and versatility of AI across various business functions. Tekpon highlights: ‘Businesses are increasingly adopting AI tools to streamline operations, make data-driven decisions, and improve customer experiences. By handling repetitive tasks and offering unique insights, AI software helps companies save time, reduce costs, and stay competitive in fast-evolving markets.’
Refact.ai is proud to be included in this recognition and will continue to build the best AI solutions to empower developers to code less and create more!
👀 Build a Website from a Sketch with Django & SQL in Seconds with Refact.ai Agent
- Explore the latest November Changelog updates in action.
- Learn how an AI Agent can build a website from a napkin sketch.
- Discover how Refact.ai Agent can help you manage your website through natural language prompts.
- Find out what’s coming next to Refact.ai Agent.
Whether you’re just starting your coding journey or have years of experience, there’s something valuable for you! Join our Discord Community to connect with like-minded developers eager to build the future of programming!
🚀 Get 1-month Free Access to the Refact.ai Pro plan
With the Refact.ai Pro plan, you get:
- Chat models that understand your codebase: Use GPT-4 Turbo, Claude 3.5 Sonnet, Groq-Llama-3.1, and more—all integrated directly into your IDE.
- AI built for your repo: Get accurate, context-aware auto-completions powered by RAG (retrieval-augmented generation) technology, tailored to specifically understand your codebase.
- All-in-one assistant for your code: Refact.ai can find and fix bugs, review code, explain snippets, refactor, and even generate new code—all with the power of advanced AI models.
Use promo code NOVEMBER to get your first month free!