Models

Deploy models on Adaptive Engine

Models power your AI applications. Add a model to a project to deploy it, then make inference requests.

SDK

Deploy a model

Add a model to your project, then deploy it:

adaptive.models.add_to_project(model="llama-3.1-8b-instruct")
adaptive.models.deploy(model="llama-3.1-8b-instruct", wait=True)

# Or use attach() to do both in one call
adaptive.models.attach(model="llama-3.1-8b-instruct", wait=True)

Parameter	Type	Required	Description
model	str	Yes	Model key from the registry
wait	bool	No	Block until model is online (default: False)
make_default	bool	No	Set as default model for the project

The model becomes available within a few minutes. Adaptive supports most transformer-based models including Llama, Qwen, Gemma, Mistral, and DeepSeek. See Integrations for proprietary models.

Multimodal models (suffixed -multimodal) include a vision encoder that is automatically spawned alongside the decoder. The vision encoder is frozen during training. See Multimodal StringThread for supported models and image handling.

See SDK Reference for all model methods.

UI

Deploy a model

Navigate to your project and open the Models tab. Click Add Model and select from the registry.

Fine-tuned adapter models appear indented under their base model.

Multimodal models (suffixed -multimodal) include a vision encoder that is automatically spawned alongside the decoder. The vision encoder is frozen during training. See Multimodal StringThread for supported models and image handling.