Clarifai Python SDK and Integrations Examples

This is a collection of examples for the clarifai-python SDK and Integrations done with Clarifai SDK. Use these examples to learn Clarifai and build your own robust and scalable AI applications.

Experience the power of Clarifai in building Computer Vision , Natual Language processsing , Generative AI applications.

Website | Schedule Demo | Signup for a Free Account | API Docs | Clarifai Community | Python SDK Docs | Examples | Colab Notebooks | Discord

Setup

• Sign up for a free account at clarifai and set your PAT key.

• You can generate PAT key in the Personal settings -> Security section

• Install the Clarifai python sdk.

• Export your PAT as an environment variable.

  export CLARIFAI_PAT={your personal access token}

• Explore and run the examples in this repo.

Usage

The examples are organized into several folders based on their category. A quick example below,

from clarifai.client.dataset import Dataset
dataset = Dataset(user_id="user_id", app_id="app_id", dataset_id="dataset_id")
dataset.upload_dataset(task="text_clf", split="train", module_dir="path_to_imdb_reviews_module")

Compute Orchestration and Agentic Examples

Function	Link	Description	Link to Docs
Basics	Basics	Basic Functionalities (create, list, patch, delete) of SDK App, Dataset, Input & Model Classes
Compute Orchestration	CRUD Operations	Basic Functionalities (create, list, get, delete) of Compute Orchestration Classes - ComputeCluster, Nodepool, Deployment
Model Predict	Predict	Call predict of any model in a Python native way with our SDK. See many more examples here	Docs for Python/Javascript
	Using OpenAI Client	Use the openAI client to call openAI-compatible models in Clarifai. See many more examples here.	Docs for Pyhon/Typescript
	Using OpenAI Client on Node.js	Use the OpenAI node client to call OpenAI-compatible models in Clarifai. See many more examples here	Docs for Typescript
	Using Vercel AI SDK on Node.js	Use the Vercel AI SDK with OpenAI provider to call OpenAI-compatible models in Clarifai. See many more examples here	Docs for Typescript
	Using LiteLLM	Use Litellm to call openAI-compatible models in Clarifai. See many more examples here
	Legacy Models	Call predict of any model not uploaded to compute orchestration with our SDK	Docs
Model Upload	Model Upload	Upload custom models, MCP tools, or any python function you want. See our new runner examples repo for compute orchestration with many examples covered!
MCP Tools	MCP Tool Examples	Upload custom MCP tools and have them fully hosted in Clarifai to use with any MCP client.
Agent Toolkits	CrewAI Examples	Build agents with CrewAI toolkits on top of Clarifai LLMs and MCP tools.	Video
	Google ADK	Create agetns with Google ADK leveraging LLMs and tools powered by Clarifai
	LLM + MCP Examples	Simple python native examples of building agents covering function calls, JSON parsing and more

SDK Notebooks

Function	Notebook	Description	Open in Colab
Basics	Basics	Basic Functionalities (create, list, patch, delete) of SDK App, Dataset, Input & Model Classes
Compute Orchestration	CRUD Operations	Basic Functionalities (create, list, get, delete) of Compute Orchestration Classes - ComputeCluster, Nodepool, Deployment
Concepts	Concept Management	Basic Functionalities (add, search, delete) of Concept Relations
Datasets	Basics	Basic Functionalities of Dataset & Input Class in SDK
	Input Upload	Upload Functionalities of SDK with different kinds of data (image, text, audio, video ) and annotations (classes, bbox, etc) using Input Class
	Dataset Upload	Upload Functionalities of SDK with different of dataset annotation formats (Clarifai, Cifar10, VOC, etc.) using Dataset Class
	Dataset Export	Export Functionalities of SDK to different of dataset annotation formats (Clarifai, Cifar10, VOC, etc.) using Dataset Class
Model Predict	Model Predict	Prediction Functionalities of SDK Model Class for different type of input data
Workflows	Create Workflow	Different kinds of Workflow Creation examples using SDK
	Patch Workflow	Modifying a workflow with patch operations using SDK Workflow Class
	Export Workflow	Exporting Workflow config and create a modified workflow using SDK Workflow Class
Model Training	Image Classification Training	Model Train demo for Visual-Classifier model type with MMClassification_EfficientNet Template
	Text Classification Training	Model Train demo for Text-Classifier model type with HF_GPTNeo_125m_lora template
	Image Detection Training	Model Train demo for Visual-Detector model type with MMDetection template
	Image Segmentation Training	Model Train demo for Visual-Segmenter model type with MMDetection template
	Transfer Learn Training	Model Train demo for Embedded-Classifier model type using Transfer-Learning
Model Evaluation	Embedding Classifier	Model Eval demo for Embedded-Classifier model type using SDK
	Text Classifier	Model Eval demo for Text-Classifier model type using SDK
	Visual Classifier	Model Eval demo for Visual-Classifier model type using SDK
	Visual Detector	Model Eval demo for Visual-Detector model type using SDK
Search	Vector Search	Introductory guide to setting up a Cross-Modal Search system using SDK
RAG	RAG	RAG setup and chat with the RAG interface.

CLI Notebooks

Function	Notebook	Description	Open in Colab
Compute Orchestration	Compute Orchestration	Basic functionalities of Compute Orchestration using CLI
Model	Model Predict	Model Prediction using CLI

Integrations Notebooks

Integration	Function	Notebook	Open in Colab
Langchain	Chains	Prompt Templates and Chains
		Retrieval QA Chain
		Router Chain
		PostgreSQL LLM
	Agents	Conversational Agent
		ReAct Docstore Agent
DeepEval	LLM Evaluation	LLM Evaluation
Unstructured.io		Github Data Ingestion
		S3 Data Ingestion
		DropBox Data Ingestion

Data Utils Notebooks

Data Util	Function	Notebook	Open in Colab
Image	Image Annotation	annotation loader
Multimodal	Data Ingestion	Ready to Use Pipelines
	Data Ingestion	Multimodal Ingestion
	Data Ingestion	Advanced Multimodal Ingestion with summarizer

Note

Although these scripts are run on your local machine, they'll communicate with Clarifai and run in our cloud on demand.

Examples provide a guided tour through Clarifai's concepts and capabilities. contains uncategorized, miscellaneous examples.