Airgapped & Self-Hosted Deployments

Jazz runs fully offline on a self-hosted server when paired with a local inference server such as Ollama or llama.cpp. Local providers need no API key, telemetry is written to local disk only, and offline mode turns off every outbound request Jazz would otherwise make on its own.

Quick setup (Ollama)

1. Install Ollama on the server and pull a tool-capable model:

   ollama pull qwen3

2. Set the environment for the Jazz process:

   export JAZZ_OFFLINE=1
   export OLLAMA_BASE_URL=http://localhost:11434

   Point OLLAMA_BASE_URL at the machine running Ollama if it lives elsewhere on the internal network. Alternatively use the config file:

   {
     "llm": {
       "ollama": {
         "base_url": "http://ollama.internal:11434"
       }
     }
   }

3. Create an agent and chat — Jazz lists models straight from Ollama’s /api/tags endpoint, so no external catalog is needed:

   jazz agent create
   jazz chat

llama.cpp works the same way via LLAMACPP_BASE_URL (default http://localhost:8080/v1); start llama-server with --jinja for tool calling.

What JAZZ_OFFLINE does

With JAZZ_OFFLINE=1 (or true), Jazz never initiates network requests on its own:

• No update check — the periodic npm registry version check is skipped (equivalent to JAZZ_DISABLE_UPDATE_CHECK=1).
• No models.dev fetch — the model catalog (used for cloud-provider model lists and metadata enrichment like context windows and pricing) is not fetched. Jazz uses the on-disk snapshot at ~/.jazz/cache/models-dev.json if one exists from a previous online run, and otherwise falls back to provider-reported metadata and defaults.
• No persona marketplace fetch — jazz persona browse reads the snapshot at ~/.jazz/cache/persona-registry.json from a previous online run, and errors if there is none. Point JAZZ_PERSONA_REGISTRY_URL at an internal catalog to browse and install inside the airgap.

Inference traffic still goes to whatever provider each agent is configured with — in an airgapped deployment that should be a local provider (ollama or llamacpp).

Model catalog options

Ollama and llama.cpp agents work with no catalog at all: model lists come from the local server, context windows and tool support are detected from Ollama’s /api/show (or llama.cpp’s /props).

If you want catalog metadata (e.g. pricing display for cloud models) inside the airgap, either:

• Seed the snapshot: run Jazz once with network access (or copy ~/.jazz/cache/models-dev.json from another machine). Every successful catalog fetch refreshes this snapshot, and offline mode reads it automatically.
• Mirror internally: host a copy of https://models.dev/api.json on your network and set JAZZ_MODELS_DEV_URL=http://mirror.internal/api.json (leave JAZZ_OFFLINE unset so Jazz fetches from the mirror).

Other network surfaces to know about

• Web tools: the web_search tool requires a configured search provider API key and will simply error without one; web_fetch and http_request reach whatever URL the agent targets — inside an airgap they can still hit internal services, which is often desirable. Network enforcement should ultimately live at the firewall.
• MCP servers: stdio servers run locally; HTTP servers connect to the URL you configure.
• Telemetry: local JSON files under ~/.jazz/telemetry — never sent anywhere.

Environment variable reference