No subscriptions, provider lock-in, or downtime. Get the best prices for every request.
$2.6
OpenPixels avg.
vs
$25
fal, replicate
per 1k Flux.1 [dev] images, 1024x576, 28 steps, on day of publication
99.9%
OpenPixels avg.
vs
98.2%
runware
past 24h uptime for Flux.1 [dev] endpoint, on day of publication
2.37s
via OpenPixels
vs
2.32s
direct
p50 latency for Flux.1 [dev] images, on day of publication
// npm install openpixels
import { OpenPixels } from "openpixels"
const client = new OpenPixels({
apiKey: "sk-op-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",
})
const result = await client.run({
model: "flux-dev",
prompt: "a cat",
})
For now we support Flux.1 [dev], Flux.1 [schnell], and Flux1.1 [pro]. We expect to add many more models soon.
We route your requests to the cheapest provider that can serve them at any given time. If they're offline, we fall back to the more reliable (and expensive) providers.
No. We run a globally distributed group of edge workers which process and route your requests.
No. We love and are inspired by their service, but they're explicitly 'a unified interface for LLMs' and we needed a router for image, video, and audio models.
Yes. You can add custom keys for each provider in your settings.
No.
Loading chart data...