Don't rawdog your prompts*

Latitude is the open-source prompt engineering platform to build, evaluate, and refine your prompts with AI

Prompt management system

Create and iterate prompts in the platform, integrate easily using our SDKs or the API
Collaborative
Build and iterate prompts with your team from an easy-to-use interface
Dynamic Prompts
Add parameters or custom logic to your prompts to support advanced use cases
Reusable Snippets
Reference shared guidelines or context from any prompt and avoid repetition
Prompt Chains
Set up chains from a single prompt document, no coding required
Version Control
Work in drafts until you're ready to go live, versions are automatically tracked in logs
API + SDKs
Integrate easily into your code with our SDKs or call our API directly

AI that builds itself

AI features to let LLMs do what they do best
You're probably using ChatGPT already for this, it's the same but without the copy and paste
Latitude uses your evaluation results to suggest improvements for your prompts automatically
Generate synthetic datasets for evaluations easily by describing the parameters you need

Evaluations

Run evaluations to check your prompts for hallucinations, accuracy, compliance, conciseness, or any other relevant metric
Run in batch
Evaluate prompt iterations in batch using production or synthetic logs
Auto-run
Monitor the performance of your prompts on a rolling basis
Human evaluations
Invite collaborators to evaluate outputs and double-check LLM evals
Template library
Latitude comes with over 20 evaluation templates ready to use or customize

Observability on autopilot

Every time a prompt runs, we automatically log the entire context, the output, and other metadata relevant for evaluations and debugging

Get started for free

You can self-host or use our Hobby tier which includes up to 40k prompt and evaluation runs every month for free

Copyright © 2024 Latitude
Copyright © 2024 Latitude