Perf Platform Documentation
Welcome to Perf - the intelligent AI runtime orchestrator that optimizes your LLM applications for cost, quality, and reliability.What is Perf?
Perf is a production-grade AI infrastructure layer that sits between your application and LLM providers (OpenAI, Anthropic, and more). We automatically route requests to the optimal model based on your task, budget, and quality requirements - while continuously learning from millions of inferences to improve over time.Why Perf?
The Problem
Building production LLM applications is complex:- Cost Uncertainty: Model costs vary 30x between providers and models
- Quality Variance: Different models excel at different tasks
- Reliability Issues: Providers experience downtime and rate limits
- Manual Optimization: Teams spend weeks tuning model selection
- Hidden Failures: Poor outputs slip through without validation
The Solution
Perf provides:- Intelligent Routing: Automatically select the best model for each request
- Cost Control: Enforce budgets and automatically optimize spend
- Quality Validation: Detect and retry poor outputs before they reach users
- Zero Downtime: Automatic failover across providers
- Learning System: Performance improves as you use it
Key Features
Smart Model Routing
- Automatic task classification (extraction, reasoning, code generation, etc.)
- Complexity-aware model selection
- Per-customer preference learning
- Real-time provider health monitoring
Cost Optimization
- Per-request budget constraints
- Automatic model downgrade when needed
- Up to 60% cost savings vs GPT-4o
- Transparent per-call billing
Quality Assurance
- Output validation (format, quality, completeness)
- Automatic retry logic
- Intelligent fallback escalation
- User feedback integration
Enterprise Observability
- Real-time performance dashboards
- Detailed call logs and traces
- Custom metrics and alerts
- ROI tracking and reporting
Production-Ready
- 99.9% uptime SLA
- SOC 2 Type II compliant
- Role-based access control
- Audit logs and compliance reporting
Quick Start
1. Get Your API Key
Sign up at withperf.pro and generate your API key from the dashboard.2. Make Your First Request
3. View Analytics
Visit your dashboard to see:- Request volume and patterns
- Cost savings achieved
- Model performance breakdown
- Quality metrics
Documentation Structure
Getting Started
API Reference
Platform Guide
Advanced Usage
Resources
Support
- Documentation: docs.withperf.pro
- Email: [email protected]
- Status Page: status.withperf.pro
- Community: Discord