Platform
Inference
Call models from several families through one OpenAI-compatible endpoint, with fallback when a provider fails.
https://api.launchad.si/v1/chat/completions -d '{"model":"launchad/default","messages":[...]}'Keep your client. Change the base URL and the key; the OpenAI SDK works as is.
Automatic fallback. A failing upstream is skipped and the next model answers.
Know who answered. Model, provider and limits come back as response headers.
Routing
One endpoint,
the right model
Pick a model by id or use an alias. If a provider times out or errors, LaunChad moves down the chain, skips upstreams that keep failing, and tells you which model answered.
Aliases launchad/default, fast, reasoning.
Circuit breakers Failing upstreams are skipped until they recover.
Streaming Server-sent events passed straight through.
HTTP/2 200x-request-id: req_8f2c41d09ab3e7x-launchad-model: glm-5.3-flashx-launchad-provider: launchadx-ratelimit-remaining: 57
Output
Tools, JSON,
and streaming
Tool calls and structured JSON output pass through to the model. Usage is counted per key, and an optional response cache serves identical deterministic requests without another model call.
Tool calling tools and tool_choice passed through.
JSON output response_format passed through.
Response cache Opt in per request with x-launchad-cache: on.
data: {"choices":[{"delta":{"content":"Hel"}}]}data: {"choices":[{"delta":{"content":"lo"}}]}data: {"usage":{"prompt_tokens":12,"completion_tokens":2}}data: [DONE]
Also included
The details
that matter
Embeddings
Through any OpenAI-compatible provider you connect.
Your provider keys
Bring OpenAI-compatible keys; stored encrypted.
Per-key limits
Requests per minute and monthly tokens.
Usage log
Model, tokens, latency, status. No prompt text.
Request lookup
GET /v1/generations/{id} for any past call.
Idempotency
Safe retries with an Idempotency-Key header.
Related
launchad · getting started
# 1. point your OpenAI client at LaunChad
export OPENAI_BASE_URL=https://api.launchad.si/v1
export OPENAI_API_KEY=$LAUNCHAD_KEY
# 2. call any model, or start an agent run
curl $OPENAI_BASE_URL/agents/researcher/runs \
-H "Authorization: Bearer $OPENAI_API_KEY" \
-d '{"message":"What changed on our competitors' pricing pages?"}'Get started
Two lines of config.
The rest is built.
Keep your OpenAI client, change the base URL and the key. Limits, logs and guardrails are on from the first request.