# NVIDIA NIM

> Catalog of hosted AI inference microservices with free API endpoints.

NVIDIA NIM hosts optimized inference microservices for dozens of leading open-weight models, including Nemotron, DeepSeek, and Llama variants. Developers can generate free API keys to prototype applications with generous rate limits without running local GPU hardware. It also offers agentic skills, workflow blueprints, and exportable containers for production deployment.

Website: https://build.nvidia.com/

## What it is for

- Power coding assistants and CLI tools via free hosted LLM API endpoints
- Prototype AI applications across different open-weight reasoning and text models
- Integrate specialized translation and embedding microservices into custom workflows
- Test agent workflows using pre-built agentic skills and blueprints

## Features

- Hosted cloud inference endpoints for open-weight foundation models
- Free API key generation with generous rate limits for prototyping
- Support for reasoning, multimodal, coding, and mixture-of-experts models
- Pre-built agentic skills and reference application blueprints
- NIM containers exportable to private cloud or local DGX infrastructure
- Standardized API schemas compatible with popular developer tooling

## Pricing

Freemium. Free tier with rate limits for testing and prototyping; paid enterprise deployment and infrastructure partnerships.

## Details

- Type: Web app
- Category: [Development](https://xiry.app/topics/development.md)
- Open source: no
- Platforms: Web, API
- For: Developers and AI engineers looking for free or low-friction hosted model APIs and microservices.
- Alternative to: OpenRouter, Groq, Together AI, Hugging Face Inference Endpoints
- Tags: apis, nvidia, nim, llm, inference, free tier, open source models

## Why it was saved

- [@nick_saraev, 31 July 2026](https://www.instagram.com/reel/Dbdq7RiPOKc/): Provides fast, free access to popular foundational models (including Nemotron and translation models) with an allowance of up to 40 requests per minute.
- [@tpadumke, 26 April 2026](https://www.instagram.com/reel/DXmWHzIExeq/): Provides 40 free LLM API calls per minute without requiring paid subscriptions, suitable as an API backend for coding assistants.

## Compare

- [Alternatives to Together AI](https://xiry.app/alternatives/together-ai.md)
- [Alternatives to OpenRouter](https://xiry.app/alternatives/openrouter.md)

## Related

- [Bytez](https://xiry.app/resources/bytez): Unified serverless API to discover and run open and closed AI models (Web app, Freemium)
- [FreeLLMAPI](https://xiry.app/resources/freellmapi): Route multiple free-tier LLM providers through a single OpenAI-compatible endpoint (GitHub repo, Freemium)
- [OpenRouter](https://xiry.app/resources/openrouter): Unified API gateway to access hundreds of AI models across providers (Web app, Freemium)
- [Firecrawl](https://xiry.app/resources/firecrawl-1984): Turn websites and documents into clean, LLM-ready markdown or structured data. (Website, Freemium)
- [json-render](https://xiry.app/resources/json-render): Generative UI framework rendering LLM JSON streams into predefined components safely (Web app, Open source)
- [awesome-llm-apps](https://xiry.app/resources/awesome-llm-apps): Open-source collection of runnable AI agent, skill, and RAG templates (GitHub repo, Open source)

Updated 3 October 2026. Page: https://xiry.app/resources/nvidia-nim

Make your own library: Xiry takes your saved reels, tweets, Reddit bookmarks and Watch Later videos, makes them searchable and pulls the useful stuff out, like this library. It's invite-only for now; join the waitlist at https://xiry.app/#waitlist.