diff --git a/docs/reference/inference-profiles.md b/docs/reference/inference-profiles.md index 86d246313..f1c1a4f49 100644 --- a/docs/reference/inference-profiles.md +++ b/docs/reference/inference-profiles.md @@ -64,13 +64,10 @@ NemoClaw validates the selected provider and model before it creates the sandbox If validation fails, the wizard does not continue to sandbox creation. -## Local Providers +## Local Ollama -Local providers use the same routed `inference.local` pattern, but the upstream runtime runs on the host rather than in the cloud. - -- Local Ollama -- Local NVIDIA NIM -- Local vLLM +Local Ollama is available in the standard onboarding flow when Ollama is installed or running on the host. +It uses the same routed `inference.local` pattern, but the upstream runtime runs locally instead of in the cloud. Ollama gets additional onboarding help: @@ -79,6 +76,13 @@ Ollama gets additional onboarding help: - it warms the model - it validates the model before continuing +## Experimental Local Providers + +The following local providers require `NEMOCLAW_EXPERIMENTAL=1`: + +- Local NVIDIA NIM (requires a NIM-capable GPU) +- Local vLLM (must already be running on `localhost:8000`) + ## Runtime Switching For runtime switching guidance, refer to [Switch Inference Models](../inference/switch-inference-providers.md). diff --git a/spark-install.md b/spark-install.md index c5f2fd0a2..a976b7192 100644 --- a/spark-install.md +++ b/spark-install.md @@ -7,7 +7,7 @@ - **Docker** (pre-installed, v28.x) - **Node.js 22** (installed by the install.sh) - **OpenShell CLI** (installed via the Quick Start steps below) -- **NVIDIA API Key** from [build.nvidia.com](https://build.nvidia.com) — prompted on first run +- **API key** for your chosen inference provider. The onboarding wizard prompts for provider and key during setup. For example, you need to provide an NVIDIA API key from [build.nvidia.com](https://build.nvidia.com) for NVIDIA Endpoints, or an OpenAI, Anthropic, or Gemini key for those corresponding providers. ## Quick Start