Langtail Roadmap, Feedback and Changelog

First of all- LOVE langtail. Right now, when running tests, it only includes the system prompt. This means anyone using one-shot or few-shot learning cannot effectively use the entire testing module. We just need a simple checkbox in the test to include the message history along with the prompt. Great work!! Thank you and hope to be a long term user.

Addidionaly, it would be useful to duplicate prompt to a new project directly

add multi model support into the prompt playground so that you can see responses from multiple models in parallel and compare them easily

Enable integration and support for Anthropic models (specifically Claude) within the application.

Develop a Python client library for Langtail, enabling seamless integration and usage of Langtail functionalities in Python applications.

There are many APIs available that are compatible with openai standard. Like fireworks.ai , togehter, claude etc. It would be great if we can add them as providers ourselves by changing the URLs

Hey, we are building an app that may send private data via LangTail. We want to make sure we dont save logs for the production deploy. We still want logs when we are testing the prompts or using other deployment tracks.

https://platform.openai.com/docs/api-reference/batch/create - this can be used to run Langtail tests cheaper, it would just take up to 24 hours to process

I store longs of my API calls to openAI, I want to be able to just copy the log and import it to Langtail so everything will be populated, including the tools and system messages.