Skip to content
Relace logo
AI & Machine LearningPrivateInferenceCompare

Relace

Relace sells small, fast models and tools for coding agents through one API: Fast Apply (merges an agent's edit into a file), Fast Agentic Search (finds the…

Relace sells small, fast models and tools for coding agents through one API: Fast Apply (merges an agent's edit into a file), Fast Agentic Search (finds the relevant code in a repo), Compact (rewrites a long agent conversation down to the parts that matter), Repos (git-style source control and sandboxes for agents), and hosted open-weight models (DeepSeek, GLM, Kimi) with automatic prefix caching. Jacq is their own macOS coding app. Relation to our focus: Compact is built around the KV cache of the big model, since it runs when the provider's cache has expired and the whole prefix would be rewritten anyway, and the hosted models bill cached input tokens at a lower rate.

Orbit · who it trades with and competes with
no links read yet
Focus · No beat
Relace
private company
No published house view
TholoBbrief←→step through the orbitESCcloseclick a company to open ittap a company to read it, tap again to open it