Langtail is an end-to-end platform for building LLM applications. It offers collaborative workspaces, a no-code playground for prompt engineering, comprehensive logging and analytics, systematic prompt testing, and streamlined deployment infrastructure. By bringing together these capabilities, Langtail enables companies to rapidly experiment with LLMs, collaborate across teams, refine prompts, and deploy production-ready applications with unprecedented speed and confidence.

LLM testing made easy with a spreadsheet-like interface. Score tests with natural language, pattern matching, or code. Optimize LLM apps by experimenting with models, parameters, and prompts. Gain insights from test results and analytics.

👋 Hi Product Hunt community! I’m Petr, co-founder and CEO of Langtail. We launched Langtail into public beta six months ago, and today we're excited to introduce Langtail 1.0 — a major step forward for building and testing AI apps. Taking an LLM app from an exciting demo to a dependable product is hard work. Over the past year, we’ve heard a lot about the challenges: LLMs are unpredictable, prompt iteration is a mess, and traditional testing tools fall short. Langtail 1.0 is our answer to these problems. It’s all about control, consistency, and confidence. Here’s what’s new: 🧪 Spreadsheet-Like Testing Interface: Drop your test cases in with ease, get instant feedback, and iterate quickly—like using Google Sheets, but for LLMs. 🔧 Hosted Tools: Langtail now handles function execution directly within the app, making prototyping easier by running tools without needing to mock responses or set up external infrastructure. 🛠️ Test Configurations: Compare different models side-by-side with just a few clicks, making it easy to see which one works best for your app. 🔗 Shareable AI Apps: Create shareable links to let anyone interact with your LLM—no sign-up required. Perfect for getting feedback or showcasing to colleagues. 🤖 Assistants: Introducing stateful assistants that handle memory and conversation history automatically, reducing the overhead of prompt management. These assistants can be tested, deployed as APIs, and even integrated across models with ease. ✨ Magic Buttons: We've added Magic Buttons to streamline workflows—automatically generate new test cases, adjust prompts, or implement improvements with a click. 🔥 AI Firewall: Real-time protection for your app. Stop prompt injections, denial of service attacks, and data leaks before they happen. 🌟 New Redesign & Light Theme: A complete redesign to improve usability, including a new light theme for those who prefer it. 🚀 Self-Hosting Available: Full control over your data, entirely on your infrastructure. Perfect for larger teams and enterprises. See more here https://langtail.com/blog/introd.... We’d love for you to give it a try and let us know what you think!

@petrbrzek The feature set looks very impressive, Petr! I shared the product in some communities that may find it useful. :)

Looks super interesting! The logs are a nice touch, but what got me truly hooked is the cost analytics to keep LLM cost under control. Well done!

@matt_roskovec Thanks Matt! Glad those metrics stuck out to you 📊 That's just the beginning of what you can do in Langtail though! Check out the tests to make sure that your prompts respond accurately too!

We’ve been using Langtail for a while at Deepnote. It’s helped us improve the reliability of our text-to-SQL system, and let us quickly prototype potential new behaviors for Deepnote AI. The Langtail team’s been super responsive to our feedback, which is great. Excited for what comes next!