Skip to content
Usefully

One API key for AI models and the live web

Call AI models through one OpenAI- and Anthropic-compatible endpoint, and search the web with the same key.

  • 0% markup on tokens
  • $1 free credit, no card
Explain automatic fallback in one sentence.
served by Anthropicfirst token 420 ms
curl https://api.usefully.info/v1/chat/completions \
  -H "Authorization: Bearer $USEFULLY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "anthropic/claude-sonnet-5-5",
    "messages": [{"role": "user", "content": "Explain automatic fallback in one sentence."}]
  }'
LLM proxy

Outages become retries your users never see

Every model is served by one or more providers. Usefully tracks each one's health and moves a request to the next healthy provider when one fails, overloads or times out.

  • Health scoring and circuit breakers per provider
  • Streaming falls back before the first token
  • Choose, order or exclude providers per request
Your app
Usefully
  • Provider Aready
  • Provider Bready
  • Provider Cstandby

Your app gets one successful response. The retry happens before it sees an error.

curl https://api.usefully.info/v1/search \
  -H "Authorization: Bearer $USEFULLY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query": "latest open-weight model releases", "num_results": 5}'
Web search

Give your model the live web

Search returns titles, URLs, snippets and dates. Ask for full page text, or read any page as clean text with one more call. Ready to drop into a prompt.

Dashboard

See every request and every cent

Each call shows its model, the provider that served it, tokens, cost and latency. Usage charts break spend down by day, key and model.

  • Cost per request, down to the token
  • Failed requests are logged and never charged
  • Per-key limits and low-balance alerts
Logs · example dataLive tail

Requests, 14 days

9,412

Spend, 14 days

$41.87

Median latency

0.82 s

Spend per day

ModelCost
anthropic/claude-sonnet-5-5$0.0124
usefully/web-search$0.0080
anthropic/claude-haiku-5-5$0.0011
usefully/extract$0.0010
anthropic/claude-opus-5-5$0.0461
Works with your tools

Keep your SDK. Change two lines.

Usefully speaks the OpenAI and Anthropic APIs, so the SDKs, frameworks and editors you use already work. Point them at Usefully and use your Usefully key.

  • OpenAI SDK
  • Anthropic SDK
  • Cursor
  • Vercel AI SDK
  • LangChain
  • LlamaIndex
  • n8n
More integrations in the docs
# Most OpenAI-compatible tools and SDKs read these two variables.
export OPENAI_BASE_URL="https://api.usefully.info/v1"
export OPENAI_API_KEY="<your Usefully API key>"

# Then use any Usefully model ID, e.g.
# anthropic/claude-sonnet-5-5

Everything you need to ship, nothing you have to run

One key, one bill

Claude models today, with more providers on the way: one API key and one balance for all of them.

Automatic fallback

When a provider errors or slows down, the request moves to a healthy one before you see it.

Provider prices

Tokens cost what the provider charges. Prepaid credits, no subscription, no minimum.

Limits you control

Per-key rate limits, daily spend caps and low-balance alerts by email or webhook.

  1. 01

    Create an account

    Sign up and create an organization. Your first one gets $1 of free credit.

  2. 02

    Make a key, try it

    Create an API key and test any model in the playground.

  3. 03

    Add credits as you go

    Buy prepaid credits by card. Watch cost per request in real time.

Questions

Something else? Get in touch.

Start building in five minutes

Create an account, get $1 of free credit, and make your first request with the SDK you already use.