MakeHub

MakeHub is a universal API load balancer for LLMs built for developers and AI infrastructure teams that run production workloads across multiple model providers.

Key capabilities include:

• Smart provider routing that dynamically selects the fastest or cheapest option using real-time signals such as latency, throughput, price, and load
• A unified OpenAI-compatible endpoint to send requests to providers like OpenAI, Anthropic, Together.ai, Fireworks, DeepInfra, Azure, plus open-model hosts, without changing client code
• Instant failover and fallback that automatically retries on backup providers to improve reliability during outages or slowdowns
• Real-time benchmarks and cost tracking to support ongoing performance and spend optimization, including arbitrage across 20+ providers

Pricing: Paid, with a flat 2% fee on credit refuel and no hidden costs.

Notable for combining provider benchmarking, routing, and reliability controls behind a single OpenAI-style API, plus a related VS Code/Open VSX extension (Apache 2.0) used in forks/enhancements of Roo-Code and Cline-style coding agents.