What is Relace
Relace builds specialised models and infrastructure for coding agents, including code search, context compaction, code application, reranking and managed repositories.
Usage-based model pricing
As of August 5, 2026, relace-search costs US$1 per million input tokens and US$3 per million output tokens. relace-compact costs US$0.20 input and US$0.20 output, relace-apply-3 costs US$0.80 input and US$1.20 output, and relace-rank costs US$0.05 per million input tokens.
The hosted API includes a free testing tier with rate limits beginning at three calls per minute. Relace Search retrieves code context, Compact reduces long agent traces, Apply merges generated changes at more than 10,000 tokens per second, and Rank scores code relevance.
Enterprise deployment
Enterprise and high-volume arrangements can provide unlimited rate limits, guided onboarding, on-premises or VPC-isolated deployment and SOC 2 controls. Hosted data is encrypted in transit and at rest; private deployments keep code within the customer's controlled environment.
Official sources: Relace pricing, Relace product documentation, Relace API and rate limits.