Skip to main content
FreeWhat would your company’s AI control layer look like?Build mine
Solutions
LLM GatewayVoice GatewayDeveloper FeaturesEnterprise Features
How it worksModelsIntegrationsDocs
Book a DemoLogin

LLM Gateway

Every model.
One endpoint.

Point any OpenAI-compatible client at Hicap. Each request goes to the right model, and fails over on its own when a provider goes down.

Hicap Gateway100+ modelsSecure and fast

Access. One key and one OpenAI-compatible endpoint put every major provider behind the client you already run.

Route. Each request goes to the cheapest model that clears your quality bar. Frontier only when it is needed.

Stay up. Automatic failover keeps answers flowing when a provider rate-limits, degrades, or goes down.

See. One log line per request: the model asked for, the model that answered, and what it cost.

What the gateway does

One endpoint.
Three jobs behind it.

Access, routing, and reliability are the same request seen from three sides. Pick one.

one key

One key. Every major provider.

Keep the client you already ship. Change a model string to change providers.

gpt-5.5 → llama-4-405b-$7.30
claude-opus-4-8 → claude-sonnet-4-6-$8.30
gemini-3.1-pro → mistral-large-3-$4.90

billed $21.90 of $42.40 list

The cheapest model that does the job.

Set the policy once. Every request runs on the smallest model that clears your bar.

  1. claude-opus-4-8 · anthropic429 rate_limited
  2. anthropic marked degradeddraining traffic
  3. gemini-3.1-pro · google200 OK

answered in 441ms

Up even when a provider is not.

Hicap watches errors and latency, and switches providers when one goes down or starts lagging.

Two lines of config.

Point the base URL at api.hicap.ai and send your Hicap key. That is the whole migration.

BalancedLatency optimizedCost optimizedData residentReserved capacity first
Browse the model catalog
1curl https://api.hicap.ai/v1/chat/completions \
2  -H "api-key: $HICAP_API_KEY" \
3  -H "Content-Type: application/json" \
4  -d '{
5    "model": "gpt-5.5",
6    "messages": [
7      { "role": "user", "content": "Hello" }
8    ]
9  }'

Reserved capacity routing

Cut your AI spend by up to 30%.

Hicap routes your traffic onto reserved capacity at a fixed price and overflows to on-demand when demand spikes. Same models, same code, a smaller bill.

  1. 01

    Reserved capacity first

    Your baseline runs on capacity bought at a fixed price.

  2. 02

    On-demand for the burst

    Spikes spill over automatically. Nothing queues, nothing drops.

  3. 03

    One bill, up to 30% lower

    Same models, same code. Only the invoice changes.

Get startedSee what you’d save