One API for every
frontier model.
Route requests to OpenAI, Anthropic and Google through a single endpoint, with automatic failover, budgets and token-level usage analytics for your whole organization.
# Same request shape as the OpenAI API
curl https://your-domain.com/api/v1/chat/completions \
-H "Authorization: Bearer $ROUTEHAVEN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "anthropic/claude-opus-5.5",
"models": ["openai/gpt-6.1-sol"],
"messages": [{"role": "user", "content": "Hello"}],
"stream": true
}'Featured models
Provider list prices per 1M tokens. Fees depend on your plan.
Anthropic's flagship Opus model for complex coding, analysis and agentic workflows.
Near-flagship quality for coding and professional work at a lower price point. Long-context (>272K) requests are billed at higher rates by OpenAI.
Google's flagship reasoning model. Introductory price through 2026-11-30; Google lists $4 / $20 per 1M tokens from 2026-12-01.
Balanced speed and capability for everyday coding, agents and enterprise workloads.
Google's most capable Flash model for agents and enterprise workflows. Price rises to $1.50 / $7.50 per 1M tokens from 2027-01-01.
OpenAI's most efficient model for focused, high-volume tasks such as classification and extraction.
Built for production traffic
Everything a platform team needs to put LLMs in front of customers — without building a gateway in-house.
Pass a list of fallback models or providers. If an upstream returns 429/5xx or times out, the next one is tried before your client sees an error.
Use your existing OpenAI, Anthropic and Google contracts. Keys are encrypted with AES-256-GCM, bound to your tenant, and never shown again.
Monthly org and team budgets, per-key spend caps, model allow-lists, and per-key rate limits — enforced before a request leaves the gateway.
Every request is logged with prompt, completion and cached tokens, latency, time-to-first-token, cost and the full fallback chain.
Server-sent events and function calling are normalised to the OpenAI format across Anthropic Messages, Gemini and OpenAI Responses.
Role-based access (Owner, Admin, Developer, Billing, Viewer), team budgets, invitations and an audit log for every sensitive change.
Up and running in three steps
Sign up, name your organization and invite your team. Assign roles and team budgets.
Add your OpenAI, Anthropic or Google API keys (BYOK). Optionally fund a balance for platform fees.
Change the base URL and API key in any OpenAI SDK. Model names look like anthropic/claude-opus-5.5.
Compliance first
RouteHaven serves traffic with your own provider credentials (BYOK). Platform-billed access for a provider is switched on only after a written reseller or distribution agreement with that provider is on file.