Migrate from fal.ai
Drop-in SDK replacement — change one line, save 30-50%
rlay.ai is API-compatible with fal.ai. The fastest migration is: change
base_url, swap your key, keep everything else.
Python
# Before
from fal_client import submit
fal_key = "fal-..."
# After
from fal_client import submit
fal_key = "rlay_sk_..."
fal_base_url = "https://api.rlay.ai"Node
// Before
import * as fal from '@fal-ai/client';
fal.config({ credentials: 'fal-...' });
// After
import * as fal from '@fal-ai/client';
fal.config({
credentials: 'rlay_sk_...',
apiUrl: 'https://api.rlay.ai',
});OpenAI SDK
For LLM endpoints, point the OpenAI SDK at our base URL:
import OpenAI from 'openai';
const client = new OpenAI({
apiKey: 'rlay_sk_...',
baseURL: 'https://api.rlay.ai/v1',
});Endpoint mapping
| fal.ai endpoint | rlay endpoint |
|---|---|
fal-ai/flux-pro | /v1/images (model: flux-2) |
fal-ai/veo-3 | /v1/videos (model: veo-3.1) |
fal-ai/kling-video | /v1/videos (model: kling-3) |
fal-ai/suno-music | /v1/music (model: suno-5.5) |
Parameter names match wherever possible. See the models page for the full mapping.
MkSaaS文档