Agent orchestration · now in GA

Ship agent workflows
with confidence

Relay is the orchestration console for AI agents in production. Route, observe, and recover every run from one surface — without rewriting your loop.

No credit card required · Deploy on your own infra

relay — session #4821
$relay run research.deep-dive --budget $2
plan · 4 steps · routed to deepseek-v4 @ $0.28
tool: web_search · 12 sources · cached hit
tool: write_report · 1,840 tokens · ok
recovered: retry ×1 on tool timeout (58ms)
done · 2m 14s · $0.61 · p95 latency 12ms
Why Relay

Everything your agents need,
nothing they don't

Four primitives cover the full lifecycle of a production agent run. Compose them, don't bolt them on.

Intelligent routing

Route every request by capability, cost, and latency. Cache hits skip the expensive path automatically — your bill drops without touching your code.

Full observability

Every tool call, token, and decision is a first-class trace. Drill from a failed run to the exact reasoning step that caused it.

Self-healing runs

Timeouts retry with backoff, failed tools degrade to fallbacks, and interrupted sessions resume from checkpoint — not from scratch.

Hard budget guards

Per-session and per-day budgets enforced at the request layer. When the budget is gone, the run stops — never silently overruns.

Any provider

DeepSeek, Anthropic, OpenAI, and your own gateway — one interface, one telemetry stream. Bring your keys, keep your data.

Human in the loop

Approvals, pauses, and handoffs where they matter. Agents run autonomously inside the envelope you define.

99.99%
Runs completed without intervention
−42%
Median token cost after routing
12ms
p95 overhead per decision
2.3M
Runs orchestrated in the last 30 days
Drop-in config

One file. Zero loop changes.

Relay wraps your existing agent — DSH, Claude Code, or your own loop — behind a thin config layer.

# relay.yaml — start with defaults, tune as you grow
session:
  budget: $5.00            # hard cap, enforced pre-request
  resume: true             # checkpoint on interrupt

routing:
  default: deepseek-v4       # cost-optimal path
  fallback: [ deepseek-v3, gpt-oss ]
  cache: read             # reuse prior turns

tools:
  timeout: 30s
  retry: 2
  degrade: true          # fall back on partial failure

telemetry:
  traces: true
  export: ./relay-runs.jsonl
Pricing

Start free. Scale when you're ready.

Starter
$0
forever, up to 100 runs/mo
  • 1 project · 2 agents
  • Basic observability
  • Community support
Start free
Enterprise
Custom
annual, self-hosted available
  • Self-hosted deployment
  • SSO & audit logs
  • Dedicated gateway
  • SLA & onboarding
Contact sales

Put your agents to work — with a leash.

Deploy in minutes, tune in days, trust from the first run.

Get started free