Getting Started
Platform Overview
Understand how Helicone solves the core challenges of building production LLM applications
Now that your requests are flowing through Helicone, let’s explore what you can do with the platform.
We fight for every penny. PTB (coming soon…) finds the absolute lowest price across providers. No markup, no games. Invisible Performance
Your app shouldn’t slow down for observability. Edge deployment keeps us under 50ms. Always. Always Online
Your app stays up, period. Providers fail, we fallback. Rate limits hit, we load balance. We don’t go down. Never Be Surprised
No shock bills. No mystery spikes. See every cost as it happens. We believe in radical transparency. Find Anything
Every request, searchable. Every error, findable. That needle in the haystack? We’ll help you find it. Built for Your Worst Day
When production breaks and everyone’s panicking, we’re rock solid. Built for when you need us most.
We built Helicone for developers with users depending on them. For the 3am outages. For the surprise bills. For finding that one broken request in millions.
How We Integrate
Helicone works through our AI Gateway - a unified API that sits between your application and LLM providers:- Single Integration - Point your OpenAI SDK to our gateway URL
- Automatic Logging - Every request and response is logged after we return it to you
- Header-Based Features - Enable capabilities like fallbacks, caching, and agent tracking via simple headers
- Zero Latency Impact - Edge deployment keeps overhead under 50ms
Currently supporting BYOK (Bring Your Own Keys) and passthrough routing. Pass-through billing (PTB) for using Helicone’s API keys is coming soon.
The Problems We Solve
Reliability Issues
Provider outages break your application. No visibility when requests fail. Manual fallback logic is complex and error-prone.
Debugging Complexity
LLM responses are non-deterministic. Multi-step AI workflows are hard to trace. Errors are difficult to reproduce.
Cost Uncertainty
Unpredictable spending across providers. No understanding of unit economics. Difficult to optimize without breaking functionality.
Prompt Management Pain
Every prompt change requires a deployment. No version control for prompts. Can’t iterate quickly based on user feedback.
Our Principles
Best Price AlwaysWe fight for every penny. PTB (coming soon…) finds the absolute lowest price across providers. No markup, no games. Invisible Performance
Your app shouldn’t slow down for observability. Edge deployment keeps us under 50ms. Always. Always Online
Your app stays up, period. Providers fail, we fallback. Rate limits hit, we load balance. We don’t go down. Never Be Surprised
No shock bills. No mystery spikes. See every cost as it happens. We believe in radical transparency. Find Anything
Every request, searchable. Every error, findable. That needle in the haystack? We’ll help you find it. Built for Your Worst Day
When production breaks and everyone’s panicking, we’re rock solid. Built for when you need us most.
Real Scenarios
Costs spiked 300% overnight
Costs spiked 300% overnight
Helicone’s user tracking and custom properties turn cost mysteries into clear insights. See exactly which users or features are driving spend with automatic cost breakdowns by user ID, feature, or any custom dimension you define. Instead of panic and guesswork, you get immediate visibility into what changed and can take targeted action.
User says AI gave wrong answer
User says AI gave wrong answer
Session tracking captures the full conversation context so you can see exactly what led to the wrong answer. Find the user’s complete interaction history, trace through multi-step workflows, and identify the exact prompt or step that failed. With prompt versioning, you can fix and deploy the correction instantly without touching code.
OpenAI is down
OpenAI is down
Automatic fallback chains keep your app running when providers fail. Configure GPT-4o on OpenAI → Vertex → Bedrock sequences that trigger instantly when requests fail or hit rate limits. Your users get the same model through a different provider, your app stays online, and you maintain full observability throughout the outage.
AI agent workflow is broken
AI agent workflow is broken
Session trees show you exactly how complex AI workflows unfold across multiple LLM calls. When multi-step agents fail, trace the entire sequence to pinpoint where it broke - whether it’s hitting token limits, using wrong context, or failing prompt logic. See the full chain of reasoning that led to the failure and fix the root cause.
Start Exploring Features
AI Gateway
Use 100+ models through one unified API with automatic fallbacks
Session Tracking
Debug complex AI agents or AI workflows
Prompt Management
Deploy prompts without code changes
Cost Tracking
Track cost and understand the unit economics of your LLM applications
We built Helicone for developers with users depending on them. For the 3am outages. For the surprise bills. For finding that one broken request in millions.