Skip to main content
SambaNova AI Starter Kits are open-source examples designed to bootstrap application development for common AI use cases. They provide Python code with concepts that can be applied to any programming language. Start with the demos below or view all kits at the GitHub repository.
To run these examples, obtain your SambaNova API key from the API keys and URLs page.

​
Quickstart

A comprehensive guide to help you get started with the platform.

GitHub

​
Enterprise knowledge retrieval

Document Q&A on multiple file formats including PDF, TXT, and DOC. Demonstrates a retrieval augmented generation (RAG) semantic search workflow.

GitHub

Cloud Demo

​
Multimodal retriever

Unlock insights from complex PDFs and images. Demonstrates advanced retrieval and answer generation that combines both visual and textual data.

GitHub

Cloud Demo

​
Benchmarking

Compare model performance by evaluating outputs, Time to First Token, End-to-End Latency, Throughput, and more with configuration options in a chat interface.

GitHub

Cloud Demo

​
Function calling

Enhance your AI applications with function calling capabilities. Demonstrates how to define and invoke functions from natural language inputs.

GitHub

Cloud Demo

​
Search assistant

Expand your application’s knowledge with semantic search and prompt construction strategies. Includes configurable integrations with multiple SERP APIs.

GitHub

Cloud Demo

​
Financial assistant

Use large language models to generate complex responses. Demonstrates how to employ multiple agents in a chain to focus and decompose queries, generate multi-step answers, and check accuracy.

GitHub

​
Agents

A multi-agent AI system that provides deep research, automated artifact production, and advanced data science capabilities. Leverages a team of specialized agents to execute complex tasks.

GitHub

Cloud Demo

​
Agent Tube

Transform video content with AI-powered analysis, transcription, and insights.

Cloud Demo

​
Voice assistant

A voice agent using LiveKit for real-time audio. Runs an STT/LLM pipeline with low-latency inference and Model Context Protocol (MCP) capabilities for advanced function calling.

GitHub

Cloud Demo

​
Third-party starter kits

Examples developed by community and ecosystem partners.

​
Video2Game

Transform educational videos (lectures, tutorials, explainer videos) into interactive games.

GitHub

​
PromptFoo

Run multiple evaluation tests on LLMs using MCP with CrewAI integration.

GitHub