SambaNova integrations enable you to build and deploy AI applications faster and more efficiently. Connect seamlessly with your favorite frameworks, data sources, and infrastructure platforms to accelerate development and streamline your AI workflows.
Developer tools
Coding assistants
| Integration | Description |
|---|---|
| BlackBox AI | Dispatches one task across multiple LLMs and merges the best result. |
| Claude Code | Scriptable CLI agent built for Unix-style piping and parallel agent teams. |
| Cline | Open-source VS Code agent that asks approval before every edit or command. |
| Codex | OpenAI’s CLI agent, sandboxed to a bounded read/write/execute scope. |
| Continue | Model-agnostic, open-source code assistant for VS Code, JetBrains, and Neovim. |
| Cursor | AI-native editor that predicts your next edit across the whole project. |
| KiloCode | Orchestrator mode splits a task across specialist agent modes. |
| OpenCode | Go CLI agent that feeds live compiler diagnostics back after each edit. |
| OpenHands | Runs every action in an isolated sandbox for autonomous software engineering. |
| Qwen Code CLI | Alibaba’s open-source Gemini CLI fork, tuned for Qwen3-Coder. |
| VS Code AI Toolkit | Microsoft’s first-party VS Code extension for building and testing agents. |
Tool and Browser Use
| Integration | Description |
|---|---|
| Browser Use | Drives a real browser by reading the page, not hardcoded selectors. |
| Composio | Connects agents to 1,000+ external apps with managed OAuth and pre-built actions. |
| Exa | Semantic search API that indexes the web and ranks results by meaning. |
| FastMCP | Turn a Python function into an MCP tool with one decorator. |
| Google Workspace | Call SambaNova models from Apps Script inside Sheets, Docs, and Gmail. |
| Instructor | Pydantic-based structured outputs with automatic retries, across 15+ providers. |
| Linkup | Web search API built for sourced, verifiable answers, not raw keyword matches. |
Frameworks
LLM frameworks
| Integration | Description |
|---|---|
| AI Suite | Switch LLM providers by changing a string, from Andrew Ng’s team. |
| Haystack | deepset’s typed, inspectable pipeline framework for production RAG. |
| Hugging Face | Hub and library for 630,000+ open models across PyTorch, TensorFlow, and JAX. |
| LangChain | Compose prompts, models, and parsers with pipe-operator chains. |
| LiteLLM | Self-hosted proxy giving one OpenAI-compatible API to 100+ LLM providers. |
| LlamaIndex | Data framework optimized for connecting LLMs to your documents via RAG. |
| Llama Stack | Meta’s standardized API layer for inference, RAG, agents, and safety together. |
| OpenRouter BYOK | Bring your own provider keys through OpenRouter’s model router. |
| Oumi | Train, fine-tune, and deploy models end-to-end from one YAML config. |
| Vercel | TypeScript AI SDK for building and streaming AI apps on Next.js and React. |
Agent building and orchestration
| Integration | Description |
|---|---|
| ADK | Google’s Agent Development Kit, with agent-to-agent interoperability. |
| Agent Zero | Autonomous personal assistant that runs on its own computer with persistent memory. |
| Agno | Lightweight, model-agnostic agent framework with microsecond agent startup. |
| AutoGen | Microsoft’s framework where agents write, run, and self-debug code together. |
| Camel AI | Research framework where role-playing agent pairs converse to complete a task. |
| CrewAI | Role-based crews of agents, coordinated sequentially or hierarchically. |
| LangGraph | Graph-based orchestration with loops, branching, and built-in checkpointing. |
| OpenClaw | Local AI assistant that replies inside WhatsApp, Slack, and Telegram. |
| Semantic Kernel | Microsoft’s SDK for wrapping your APIs as callable LLM plugins. |
Interfaces and low-code
LLM Interfaces and Gateways
| Integration | Description |
|---|---|
| Gradio | Wrap any Python function into a shareable web demo, no JS required. |
| Open WebUI | Self-hosted, provider-agnostic chat UI with built-in RAG and code execution. |
Low-code platforms
| Integration | Description |
|---|---|
| Dify | Full-stack LLM platform: workflows, RAG pipeline, and API publishing. |
| Flowise | LangChain.js-based drag-and-drop builder, tuned for lightweight concurrency. |
| Langflow | Python-native visual builder with full component source access, backed by DataStax. |
| Make | Cloud-only, point-and-click automation with 3,000+ pre-built app modules. |
| n8n | Open-source, self-hostable automation with native AI-agent and RAG nodes. |
Data and infrastructure
Hyperscalers
| Integration | Description |
|---|---|
| AWS Marketplace | Deploy and pay for SambaNova through your existing AWS billing relationship. |
| AWS PrivateLink | Route traffic to SambaNova over AWS’s private backbone, off the public internet. |
Orchestration platforms
| Integration | Description |
|---|---|
| DataRobot | Enterprise governance and real-time moderation for production agents. |
| Docker Compose for Agents | Define LLMs, MCP tools, and agents from any framework in one Compose file. |
Vector DB and search
Multimodal
Real-time voice
| Integration | Description |
|---|---|
| ElevenLabs | Realistic text-to-speech and voice cloning, extended into full voice agents. |
| Hume AI | Detects tone and emotion in speech to drive empathic voice responses. |
| LiveKit | Open-source WebRTC infrastructure that other voice frameworks run on top of. |
| Pipecat | Vendor-neutral Python pipeline chaining STT, LLM, and TTS in real time. |
| Vapi | API-first platform purpose-built for phone-call voice agents. |
Video Analysis
| Integration | Description |
|---|---|
| TwelveLabs | Video-native AI that searches and summarizes footage by what’s actually shown. |
Evaluation and monitoring
| Integration | Description |
|---|---|
| Inspect AI | UK AI Security Institute’s sandboxed framework for agent safety evals. |
| LM Evaluation Harness | EleutherAI’s standardized runner for hundreds of reproducible model benchmarks. |
| Weave | W&B’s tracing tool: decorate a function to log every call into an explorable tree. |