Codeflash
Make your Python code run up to 10x faster automatically
128 followers
Make your Python code run up to 10x faster automatically
128 followers
Codeflash optimizes all your Python code. It uses advanced AI and verification to automatically find the most optimized version of any Python code. Installing Codeflash as a GitHub Action or VS Code extension ensures that all new Python code your team and your agents write is always optimal and runs fast.
Interactive
Free Options
Launch Team / Built With
👋 Hey Product Hunt! I'm Saurabh, Founder and CEO of Codeflash.
We're excited to launch Codeflash after more than 2 years of work to solve a problem we're passionate about—writing code that runs fast.
Right now, while writing code we all focus on making it work correctly, but optimizing code is hard and time-consuming. It often gets skipped in favor of building new features or fixing bugs. The result? We ship code with hidden performance problems that makes software run far slower than it should. I've found so many 10x speedups where I thought, "How did this slip by?"
But we all want software that runs fast! So how can we make all code run fast at scale?
We are creating a new way with Codeflash! Codeflash discovers the fastest version of your Python code through AI-powered optimizations, verifies correctness, and automatically delivers performance gains. Now you can write code quickly and have it run fast!
You can use Codeflash by installing it as a GitHub Action that optimizes all new code in pull requests. When it finds a high-quality optimization that's useful, correct, and faster, it creates a suggestion with the code diff for you to review. Your team never ships slow code, and your project stays fast with every change.
Codeflash is also available as a VS Code/Cursor extension that lets you optimize any code in your project with just one click—right in your IDE!
Codeflash is a general-purpose optimizer that works on almost any type of Python code: Machine Learning (PyTorch, etc.), Algorithms, Numerical code (NumPy, etc.), Data processing (pandas, etc.), AI Agents, and backend systems (async code). We're trusted by leading engineering teams at vLLM (PRs), Roboflow (PRs), Unstructured.io(PRs), Langflow (PRs), Pydantic (PRs), and more who now always ship performant code. We've sped up end-to-end latencies for companies by upto 35%, for products used by millions daily.
This is our first public launch, and we'd love your thoughts, feedback, or even just a "hey 👋".
If you're curious how Codeflash can help your team, 📆 book a call or email me directly at saurabh@codeflash.ai.
Let's build a future where software always runs fast by default. 🚀
Command Center
I used an early version of this. It made one of our hot workflows 25% faster and reduced our AWS bill by 5%. This product prints money, and you'd be nuts not to use it.
@jimmykoppel That's impressive results! Glad you found great results! Would love if you could try out our VS Code extension and let us know your feedback for us, would value it a lot.
Congratulations! I hope to get my hands dirty with this tool, writing performant python has always been tricky for me.
@aseembits93 thank you, let me know what feedback you have for me as you try it out
Snowglobe
This is really cool! Performance optimization is one of those things that always gets pushed to "later" - love that you're making it automatic. Excited to see how this helps teams ship faster code without the headache. Congrats on the launch!
@alejandroesquivel I agree, thats why we want to solve this problem! Curious to hear what results you find on your codebase.
Having worked at Datadog building tracing libraries and profilers, this is a major stride towards scaling AI to optimize to the line of code. Kudos!
@priyanshi_gupta9 thank you!
Swytchcode
Brilliant idea. Congrats on the launch!
@chilarai thank you!
Workplan
@amir_shevat1 Thank you!