Last updated: 11/12/2026Last verified: 2026-11-12

ARC by Rayline

ARC by Rayline Overview: Model Routing for AI Coding Workflows

0
0

ARC is Rayline’s routing model: it chooses a model and a reasoning level for each turn of an agent or API request. You can call fixed-rate ARC Swift or ARC Sage, pay the served model’s catalog price with ARC Dynamic (0% commission), or use the Routes API to get a selection and run inference yourself. The hosted API is in beta; the ARC-1.0 weights on Hugging Face are Apache-2.0.

Paid

What is ARC by Rayline?

ARC by Rayline is a routing model that selects a model and reasoning level for each turn of an agent workflow or API request. It is relevant to AI coding setups where developers want model selection to happen dynamically rather than choosing one model for every task. Rayline offers hosted inference options, a Routes API for selection without hosted execution, and Apache-2.0-licensed ARC-1.0 weights on Hugging Face. The hosted API is currently described as being in beta.

Key Features

  • Per-turn model and reasoning selection: ARC chooses both a model and a reasoning level for individual agent turns or API requests.
  • ARC Swift and ARC Sage: The arc-swift-1.0 and arc-sage-1.0 endpoints combine routing with fixed US-dollar-per-million-token rate cards.
  • ARC Dynamic: The arc-1.0 endpoint is billed at the served model’s Rayline catalog price, with 0% commission and no extra routing fee.
  • Routes API: Returns a model selection so developers can execute inference themselves.
  • Decisions API: Answers typed questions about a state without generating text.
  • API compatibility: The hosted endpoint at https://api.rayline.ai supports Anthropic Messages and OpenAI Chat Completions interfaces.
  • Coding workflow documentation: Rayline lists integration guides for OpenCode, Hermes, OpenClaw, Claude Code, Claude Desktop, Codex, pi, and other tools.
  • Available model weights: ARC-1.0 weights are published on Hugging Face under Apache-2.0.

Best For

Developers building coding agents that need model and reasoning-effort selection at each turn.Teams comparing fixed token-rate routing with billing based on the model actually served.API developers who want routing decisions while retaining responsibility for inference execution.Developers integrating routing into tools that use Anthropic Messages or OpenAI Chat Completions interfaces.Researchers and engineers interested in evaluating the Apache-2.0-licensed ARC-1.0 weights.

Pricing

ARC hosted requests are paid from a Rayline Platform prepaid balance (a new workspace starts at zero; an unpaid request is refused with HTTP 402, and Rayline does not substitute a cheaper model). Current token rates were read from the public feed https://api.rayline.ai/v1/arc/prices on 2026-11-12 (card version 64, asOf 2026-11-11T12:30:02.290Z, USD per 1M tokens), which is the live card pointed to by https://rayline.ai/pricing and https://docs.rayline.ai/arc/pricing.md. ARC Swift 1.0: input $0.10, output $0.50, cache read $0.01, cache write $0.15. ARC Sage 1.0: $1 / $5 / $0.05 / $1.25. The Routes API and Decisions API (arc-routes-1.0 and arc-decisions-0.1): input $0.042, cache read $0.0042, with no output or cache-write charge. Contributor and zero-data-retention IDs have their own cards on the same feed. ARC Dynamic (arc-1.0) is billed at the served model’s catalog price, with 0% commission and no extra routing fee. Workshop subscription plans are a separate product.

Pros & Cons

Pros

  • Selects both the model and reasoning effort rather than handling model choice alone.
  • Offers fixed-rate and served-model catalog-price billing approaches.
  • Separates routing from inference through the Routes API for developers who want to execute requests themselves.
  • Supports two widely used API interfaces and provides documentation for multiple coding tools.
  • Makes ARC-1.0 weights available under Apache-2.0.

Cons

  • Official docs describe the hosted API as in beta, and some OpenAPI fields are marked as proposed.
  • A new workspace has no prepaid balance. Requests the workspace cannot pay are refused with HTTP 402; Rayline does not automatically switch to a cheaper model.
  • The Routes API does not yet reuse a cached prefix between requests, so every token it reads is billed at the encoded-input rate.
  • The Decisions API is limited to 30 requests per minute per user, counted separately from the Routes API.
  • Routing quality, latency and coding-task performance were not independently benchmarked for this listing.

Alternatives

OpenRouter

A multi-provider model gateway, rather than a routing model that picks a model and reasoning level for each turn.

Wayfinder Router

An open-source router you run yourself, instead of Rayline’s hosted Swift, Sage and Dynamic APIs.

Vela 2.0

Open foundation routing models, rather than Rayline’s hosted ARC API and published rate cards.

FAQ

Details

Platform

APIWeb

Features

  • ARC Swift (arc-swift-1.0) and ARC Sage (arc-sage-1.0) pick a model and reasoning effort for each request, billed on a fixed US-dollar-per-million-token rate card
  • ARC Dynamic (arc-1.0) is billed at the served model’s Rayline catalog price, with 0% commission and no extra routing fee
  • The Routes API returns a model selection for you to execute; the Decisions API answers typed questions about a state and generates no text
  • The API speaks Anthropic Messages and OpenAI Chat Completions at https://api.rayline.ai, and the docs list guides for OpenCode, Hermes, OpenClaw, Claude Code, Claude Desktop, Codex, pi and other tools

Languages

en

Known limitations

  • Current rates come from the public feed https://api.rayline.ai/v1/arc/prices (card version 64, asOf 2026-11-11T12:30:02.290Z, units are USD per 1M tokens) and are explained on https://docs.rayline.ai/arc/pricing.md, which points to rayline.ai/pricing for the live card. ARC Swift 1.0: input $0.10, output $0.50, cache read $0.01, cache write $0.15. ARC Sage 1.0: $1 / $5 / $0.05 / $1.25. Routes API and Decisions API (arc-routes-1.0 and arc-decisions-0.1): input $0.042, cache read $0.0042, no output or cache-write charge. Contributor and zero-data-retention IDs have their own cards on the same feed (Swift Contributor $0.05/$0.25/$0.005/$0.075; Sage Contributor $0.5/$2.5/$0.025/$0.625; Swift ZDR $0.05/$0.25/$0.005/$0.0625; Sage ZDR matches Sage’s standard card; Routes/Decisions Contributor $0.03 input and $0.003 cache read; Routes/Decisions ZDR $0.08 input and $0.008 cache read)
  • Hosted requests are paid from a workspace’s prepaid balance on the Rayline Platform. Docs (https://docs.rayline.ai/platform/billing.md) say a new workspace has no balance, a request the workspace cannot pay is refused with HTTP 402 and error type billing_error, and Rayline does not substitute a cheaper model. A workspace linked to a Workshop organization uses that organization’s Workshop credits instead; Workshop plans (Discover free, Build $20/month, Scale $50/month, Max $200/month, plus Enterprise via sales) are a separate product, documented at https://docs.rayline.ai/workshop/pricing/plans.md
  • Docs (https://docs.rayline.ai/llms.txt and availability) describe ARC as in beta, with some OpenAPI fields marked x-status: proposed. The public GET /v1/models list on 2026-11-12 returned only arc-swift-1.0 and arc-sage-1.0; both advertise a 1,000,000-token context window, 128,000 max output tokens, and vision, tools and reasoning. The Routes API does not yet reuse a cached prefix between requests, so every token it reads is billed at the encoded-input rate. The Decisions API rate limit is 30 requests per minute per user, counted separately from the Routes API; refused Decisions requests are not billed
  • The announcement post describes ARC as fully open-source. The weights published as rayline-ai/ARC-1.0 on Hugging Face are licensed Apache-2.0 (base model Qwen/Qwen3.5-0.8B). Raydock is documented as open source with no software fee, with hardware and cloud inference billed separately. This listing is based on the marketing site, the pricing page, the public prices API, official docs and the Hugging Face model card; an account was not created and no paid request was sent

Official pricing source

https://rayline.ai/pricing

Rate This Tool

Related Tools