Providers¶

Tarash Gateway provides a unified interface to generate videos and images across multiple AI providers. Change provider and model in your config — nothing else in your code changes.

Supported providers¶

Video & Image¶

Provider	Video	Image	Image-to-Video	Install extra
OpenAI	✅	✅	✅	`openai`
Azure OpenAI	✅	✅	✅	`openai`
Fal.ai	✅	✅	✅	`fal`
Google	✅	✅	✅	`veo3`
Runway	✅	—	✅	`runway`
Replicate	✅	—	✅	`replicate`
xAI	✅	✅	✅	`xai`
Stability AI	—	✅	—	—

Audio (TTS / STS)¶

Provider	TTS	STS	Install extra
Fal.ai	✅	—	`fal`
ElevenLabs	✅	✅	`elevenlabs`
Cartesia	✅	✅	`cartesia`
Sarvam AI	✅	—	`sarvam`
Hume AI	✅	—	`hume`

Switching providers¶

from tarash.tarash_gateway import generate_video
from tarash.tarash_gateway.models import VideoGenerationConfig, VideoGenerationRequest

request = VideoGenerationRequest(
    prompt="A cat playing piano, cinematic lighting",
    duration_seconds=4,
    aspect_ratio="16:9",
)

# Use Fal.ai
config = VideoGenerationConfig(provider="fal", model="fal-ai/veo3.1/fast", api_key="YOUR_FAL_KEY")
response = generate_video(config, request)

# Switch to Runway — same request, one line change
config = VideoGenerationConfig(provider="runway", model="gen4_turbo", api_key="YOUR_RUNWAY_KEY")
response = generate_video(config, request)

For the full list of model IDs per provider, click any provider name in the table above.

For automatic failover between providers, see the Fallback & Routing guide.