Relace
Provides purpose-built AI models and infrastructure for coding agents, enabling ultra-fast code retrieval, merging, and autonomous workflows.
The product provides purpose-built AI models and infrastructure for coding agents, enabling ultra-fast code retrieval, merging, and autonomous workflows. It solves the problem of slow, unreliable code generation by replacing general-purpose LLMs with specialized small language models. It is sold to software developers, engineering teams, and enterprises building or integrating coding agents. It differentiates by offering in-house trained models that outperform frontier LLMs on utility tasks, with integrated source control and fast apply merging at 10,000 tokens per second.
Key features
- Semantic code retrieval
- Fast merging at 10,000 tok/s
- Specialized small language models
- Source control with branching and indexing
- Lightweight push/pull operations
- Automatic codebase indexing