Rayline launches ARC, a per-turn model router claiming 90% cost cuts
testingcatalog · x · 2026-11-12
Rayline has launched ARC, a per-turn model routing service that picks the right model and reasoning level for each request, claiming routing at 3.2¢ vs 32¢ for running Opus every turn — roughly 90% cheaper.
ARC ships two modes: a Generation API where Rayline runs the selected model and returns the response, and a Routes API that sends you the selection to run on your own endpoints, with tunable exploration temperature and routing value. It integrates with llama.cpp, vLLM Semantic Router, OpenRouter, Claude Code, Cursor, Cline, LangGraph and more; weights are available on Hugging Face.
Related event: Rayline Launches ARC Model Router at 90% Lower Cost(2 posts)→
More from coding & agent
- Stanford NLP course's best homework: build your own agent harness — stanfordnlp · 2026-11-12
- Anthropic details agent memory: dreaming organizes Claude's memory, wiki+search hits 61% — RLanceMartin · 2026-11-12
- Talk preview: scaling 1,000 Fable 5.2 agents to formalize Fermat's Last Theorem — RLanceMartin · 2026-11-12
- Opinion: AI infra is becoming a coding runtime as intelligence units outgrow tokens — prateekj · 2026-11-12
- Claude Docs/Slides block external MCP clients: no public API, export only — steipete · 2026-11-12
- Anthropic explains agent memory: 'dreaming' lets Claude consolidate session transcripts into notes — rseroter · 2026-11-12