33 models · one API · failed calls are free

Images, video and code, all from one API.

Use them right here in the browser, or point the OpenAI SDK at us and change one line.

A picture made on the platform — ImagesFrom $0.0065 per run

A frame from a video made on the platform — Video

GPT-6 LunaTurn this SQL into a weekly rollupSELECT date_trunc('week', created_at)…$0.10 in · $0.50 out / 1M

What do you want to make?

All 33 models →

Make images12 models · From $0.0065 per run

Make video14 models · From $0.0112 per run

{
  "model": "…",
  "messages": [
    { "role": "user",
      "content": "…" }
  ]
}

Chat and code5 models · From $0.10 per 1M input tokens

Music and 3D2 models · From $0.07 per run

Most used this week

GPT Image 2.5 Ext

$0.0065 / call

$0.010625

No. 1 this week </>GPT-6 Luna

$0.10 in · $0.50 out / 1M

No. 2 this week

GPT Image 2

$0.0065 / call

No. 3 this week

Already on the OpenAI SDK? Change two lines.

Streaming, tool calls, image and video tasks keep working. Failed requests are not charged, and tasks settle at what they actually used.

 const client = new OpenAI({
-  baseURL: "https://api.openai.com/v1",
+  baseURL: "https://www.aisplice.store/v1",
-  apiKey: process.env.OPENAI_API_KEY,
+  apiKey: process.env.GATEWAY_API_KEY,
 });

Read the integration guide

curl https://www.aisplice.store/v1/chat/completions \
  -H "Authorization: Bearer sk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "GPT-6 Luna",
    "stream": true,
    "messages": [
      { "role": "user", "content": "Say hello." }
    ]
  }'

Failed calls are free

A failed call or a failed generation is not charged. Tasks are estimated first and settled at what they actually used.

Healthy routes first

When an upstream struggles, requests move to a healthy one. You do not write the retry.

Every call on the record

The request log shows what each call cost, its tokens and what it made. The status page is public.