Free desktop application for discovering, running, and chatting with local open-weight models.
Nvidia Build
FreemiumNVIDIA's developer hub to test foundation models, blueprints, and OpenAI-compatible NVIDIA NIM inference APIs.
Overview
What Nvidia Build does
Nvidia Build (build.nvidia.com) is NVIDIA's developer portal for exploring, benchmarking, and prototyping with NVIDIA NIM microservices across leading open and proprietary models including Nemotron, DeepSeek, Llama, and vision-language models. Developers can run interactive browser tests, call hosted endpoints via an OpenAI-compatible API, and deploy reusable agentic skills and reference blueprints.
Capabilities
Key features
- Interactive catalog of foundation, multimodal, and Nemotron models hosted on NVIDIA NIM
- Drop-in OpenAI-compatible REST endpoints for rapid API prototyping
- Curated collection of AI blueprints, agentic skills, and hardware playbooks (including NemoClaw)
- Downloadable NIM containers for self-hosted deployment under NVIDIA AI Enterprise
Fit & Workflows
Best-fit use cases
- AI engineers prototyping inference pipelines and comparing open-weight models quickly
- Enterprise developers building RAG or agent applications using reference blueprints
- Teams planning to transition from hosted test APIs to self-hosted GPU deployments
Browse matching directory filters:
Compare Alternatives
Related AI tools & alternatives
Full-stack AI app builder combining multi-agent code generation with built-in database, auth, storage, and hosting.
Agentic AI pentesting and automated DAST platform for securing web apps, SPAs, and APIs.