Public beta · OpenAI-compatible
Omniflare is a drop-in gateway to OpenAI, Anthropic, and every major LLM provider — with prepaid budgets, hard spend caps, and one consolidated invoice for your whole team.
from openai import OpenAI
client = OpenAI(
base_url="https://api.omniflare.com/v1", # the only change
api_key="omni_sk_live_...", # one key, every provider
)
response = client.chat.completions.create(
model="anthropic/claude-sonnet-5",
messages=[{"role": "user", "content": "Ship it."}],
)How it works
Keep the SDK and code you already have. Omniflare slots in between your app and the providers — enforcing your budget on the way through.
Point the OpenAI SDK — or any OpenAI-compatible client — at Omniflare. Change one line: the base URL.
We authenticate, enforce your budget and spend caps, route the request, then meter usage — all in a few milliseconds.
The call reaches OpenAI, Anthropic, Google and more. One response format back, one line on your bill.
Product
Four pillars that turn a pile of provider accounts into one gateway you can actually govern.
One key reaches every major provider. Because Omniflare speaks the OpenAI API, adopting it is a one-line change — no rewrites, no new SDKs.
Prepaid balances plus a defined overdraft, with pre-authorization on every request. Hard caps mean a runaway loop or a bad deploy can never blow past your limit.
Stop reconciling a dozen provider invoices. Omniflare consolidates every model and every provider into a single, itemized bill for the whole team.
See exactly where spend and tokens go — in real time. Break usage down by API key, model or team to catch anomalies before they reach the invoice.
Quickstart
If your code already speaks the OpenAI API, you're basically done. Point it at Omniflare and every provider is available through the same call.
One key authenticates against every provider. No per-vendor setup.
Point your existing OpenAI client at api.omniflare.com — one line.
Prefix the model with its provider, e.g. anthropic/… or openai/….
- base_url = "https://api.openai.com/v1"
+ base_url = "https://api.omniflare.com/v1"
# everything else stays exactly the samecurl https://api.omniflare.com/v1/chat/completions \
-H "Authorization: Bearer $OMNIFLARE_API_KEY" \
-d '{ "model": "openai/gpt-5", "messages": [{"role":"user","content":"Hello"}] }'Pricing
Start free and scale when you need to. Your plan is a flat platform fee — provider usage is always billed at cost from your prepaid balance.
Evaluate the gateway and ship side projects.
For solo builders shipping to production.
For teams that need real control.
For high volume and compliance needs.
All plans include access to every provider. Usage is metered per token and billed at provider cost — never marked up.
Trusted by teams building with AI
Company logos shown are placeholders.
Start free
One key, every model, and budgets that can't be blown. Create an account and make your first call in minutes — no credit card required.