scorchmark is an open-source CLI tool and MCP server that helps AI infrastructure teams detect cache-TTL waste, simulate model-swap savings, and analyze pricing drift across multiple AI providers. It provides per-agent attribution and integrates with prompt-caching workflows, supporting cost optimization and observability for large language model deployments.
scorchmark is an Other LLM eval project. It focuses on identifying and reducing unnecessary cache costs and inefficiencies in AI model usage across providers. It is built as an open-source project for AI infrastructure engineers and FinOps teams. scorchmark is open source under the MIT license. scorchmark is available on the command line and API, and it can be self-hosted.
Nas01010101 builds and maintains scorchmark, and it first shipped in 2026. Key capabilities include Cache TTL analysis, model swap simulation, and pricing drift detection. It exposes integrations via an MCP server and a public API.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Same category — not a similarity match