Awesome LLM Cost
A curated list mapping the tools, libraries and patterns for reducing the cost of running large language models in production, covering observability, routing, caching, prompt compression, quantization and provider pricing data.
Awesome LLM Cost collects open-source tools and reference material for lowering what a team spends on LLM inference, scoped explicitly to cost rather than general LLM observability, raw inference speed or prompt-engineering-for-quality work. Sections cover cost tracking and budget observability platforms such as Langfuse, Helicone and Phoenix; routing and model-selection libraries; caching and semantic caching; prompt compression; quantization, distillation and other model compression research; self-hosting starting points; batching and async serving patterns; and pricing data, calculators and benchmarks for Anthropic, OpenAI and Gemini models.
Each entry is a one-line description of what the tool does and, where relevant, whether it is self-hostable. The maintainer states every entry was verified as actively maintained and in real use as of May 2026, and the list is scoped to stay a single reference map across a set of otherwise scattered resources.
It is a link index, not a tool or library of its own.