Migration Guide

Switching to FOTOhub from another AI provider? This guide maps your existing code to FOTOhub equivalents.

From OpenAI (DALL-E / GPT)

Image Generation

OpenAI (before)

from openai import OpenAI
client = OpenAI(api_key="sk-...")

response = client.images.generate(
    model="dall-e-3",
    prompt="A sunset over mountains",
    size="1024x1024",
    n=1
)
url = response.data[0].url

FOTOhub (after)

from fotohub import FotoHub
client = FotoHub(api_key="fh_live_...")

result = client.generate_image(
    prompt="A sunset over mountains",
    model="seedream-5-0-260128",  # Better quality, lower cost
    aspect_ratio="1:1"
)
url = result["images"][0]

Chat Completions

FOTOhub is OpenAI-compatible in message format, but not in streaming — see the warning below.

OpenAI (before)

from openai import OpenAI
client = OpenAI(api_key="sk-...")

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

FOTOhub (after)

from fotohub import FotoHub
client = FotoHub(api_key="fh_live_...")

# Same message format. chat() returns a plain dict -- index it.
response = client.chat(
    messages=[{"role": "user", "content": "Hello!"}],
    model="gemini-flash",  # or "gemini-pro", "gpt-4o", "claude-sonnet"
)
print(response["choices"][0]["message"]["content"])
print(response["credits_used"])

Streaming does not carry over

/v1/ai/chat/completions accepts stream=True for drop-in compatibility and then ignores it — you always get one complete JSON body, never chat.completion.chunk frames. An OpenAI-style for chunk in stream: loop over it yields nothing.

The model list is also narrower than the catalogue: exactly gemini-flash, gemini-pro, gpt-4o and claude-sonnet. Anything else (including claude-sonnet-4.6) returns 400.

To actually stream, switch to POST /v1/ai/agent/stream, which uses its own type-keyed frames and its own model IDs. See the Streaming Guide.

From Stability AI

Stability (before)

import requests

response = requests.post(
    "https://api.stability.ai/v2beta/stable-image/generate/core",
    headers={"authorization": f"Bearer sk-..."},
    files={"none": ''},
    data={"prompt": "A lighthouse at dusk", "output_format": "png"}
)

FOTOhub (after)

from fotohub import FotoHub
client = FotoHub(api_key="fh_live_...")

# Stability tools available natively + 20+ other image models
result = client.generate_image(
    prompt="A lighthouse at dusk",
    model="seedream-5-0-260128"  # or "flux-2-pro"
)

Background Removal

# Stability API → FOTOhub (same underlying model, lower price)
result = client.stability_remove_background(image_base64)

From Replicate

Replicate (before)

import replicate

output = replicate.run(
    "black-forest-labs/flux-pro",
    input={"prompt": "A cat in space", "steps": 25}
)

FOTOhub (after)

from fotohub import FotoHub
client = FotoHub(api_key="fh_live_...")

result = client.generate_image(
    prompt="A cat in space",
    model="flux-2-pro"
)
# Same FLUX model, unified billing, no cold starts

From ElevenLabs (TTS)

# FOTOhub wraps multiple TTS providers under one interface
result = client.generate_speech(
    text="Welcome to FOTOhub!",
    model="google",       # or "elevenlabs"
    voice_id="alloy",
    language="en"
)
print(result["audio_url"])

Key Differences

Feature	Other Providers	FOTOhub
Billing	Per-provider accounts	Single wallet, credits or tokens
Models	1 provider = 1 SDK	100+ models across 10+ providers, 1 SDK
Failover	Manual	Manual — pick another model ID and retry
Rate limits	Per-model	Tier-based, shared across all
Video polling	Custom implementation	Built-in wait_for_video()
Model selection	You decide	You decide, or ask POST /v1/ai/gabriel for a recommendation

There is no automatic failover

Earlier revisions of this table promised an automatic model chain. The public API does not retry a failed generation against a different model — a provider outage surfaces as an error on that request. Build retry/fallback into your own client if you need it.

Gabriel returns a routing decision (which model to use); it does not run the generation. See Cost Optimization for the real request shape.

Migration Checklist

• [ ] Create FOTOhub account at fotohub.app
• [ ] Generate an API key in Console → Keys
• [ ] Install SDK: pip install fotohub or npm install fotohub
• [ ] Replace provider-specific API calls with FOTOhub equivalents
• [ ] Set up webhooks for async operations (video, 3D)
• [ ] Configure spending limits in Console → Billing
• [ ] Test with sandbox key (fh_test_...) before going live