Store your agents privately. See which version each enrolled computer is running, try updates on a small group first, and roll back from your browser.
Cloud coordinates releases. Your computers run the models, keep their working version when an update fails its tests, and keep running offline.
Free for 90 days. No card required.
Open your organization in Cloud and choose Fleet. Manage releases and see reported status without signing in to every host.
View enrolled hosts, their assigned channel, active agent and runtime, acceptance-test results, hardware totals, and last check-in.
Start with canary hosts before opening the rollout to the rest. Every host runs the included tests locally before activating the exact published version.
A rejected candidate pauses further activation. A host that rejects it keeps its working copy. Start a tested rollback from the same browser dashboard.
Install a Fleet-enabled NexusRun build, import the release's exact managed CPU runtime, and enroll the host for one repository and channel. Fleet uses sealed, single-model chat agents with an acceptance suite on Linux x86-64 CPU.
Hosts poll over outbound HTTPS; the browser does not run the model or need an inbound connection to your computers. Updates apply to subsequent runs. Last check-in is a report, not a live process-health check.
The trouble starts when the agent has to run on 40 machines you don't administer: a locked-down Linux box at a clinic, a line PC on a plant floor, a kiosk, a host that never touches the internet.
Every machine pulls the same versioned unit from one place, instead of a shared drive and a hope.
Your units live in your own namespace. Nobody else can list them, and pulls need a token you issued.
Export a sealed file, carry it across, and import it. The registry is optional on the far side.
Public registries aren't private, Kubernetes is too much for a kiosk, and Docker often isn't allowed on the box in the first place.
A unit packages an agent's instructions and configuration. Linked units reference an external model; sealed units include the weights. Fleet releases use sealed units so hosts receive the exact model bytes too.
nexus login nexus build ./triage-agent --seal nexus push triage-agent:1.4.0 registry.nexusrun.dev/acme/triage-agent:1.4.0
It's a standard OCI registry, so nexus push is all it takes.
nexus pull registry.nexusrun.dev/acme/triage-agent:1.4.0
nexus doctor # reports what the hardware can actually run
nexus run triage-agent:1.4.0 -p "summarize today's intake"
Artifact downloads are delivered through a CDN. For managed updates, enroll the host and use Fleet to select the exact release instead of updating references by hand.
nexus build ./triage-agent --seal -o triage.nx
nexus import triage.nx # on a prepared offline host
For a clean destination, the offline guide packages the agent, NexusRun, and a pinned runtime in a complete bundle. Carry it by USB or your approved transfer process.
| Stays on your machines | NexusRun Cloud hosts |
|---|---|
| Model weights, unless you choose to push a sealed unit | Unit artifacts (OCI) you push |
| Live inference inputs, generated answers, and local logs | Your org, its members and access tokens (stored hashed) |
| Secret values, in a local encrypted store | Storage and bandwidth usage for billing |
| Inference: llama.cpp or Ollama on your hardware | Fleet assignments, release history, and last check-ins |
| The NexusRun runtime (Apache 2.0, free forever) | Reported hardware totals and aggregate acceptance results |
There is no hosted inference and no per-token bill. If NexusRun Cloud went down tomorrow, units already on your machines would keep running.
Agent instructions and acceptance-test fixtures travel inside the artifact you publish. Use synthetic test inputs; a sealed Fleet unit also uploads its model weights to your private registry.
Private artifact delivery and Fleet management, with plan limits for hosts, storage, and bandwidth. No per-token inference charges.
| Free trial | Developer | Team | Enterprise | |
|---|---|---|---|---|
| Price | $0 for 90 days, no card | $19/mo | $149/mo | Custom |
| Storage included | 5 GB | 10 GB | 100 GB | Negotiated |
| Pull bandwidth / mo | 20 GB | 100 GB | 500 GB | Committed |
| Seats | 1 | 2 | 5 | Custom |
| Enrolled hosts | 10 | 25 | 100 | Custom |
| Private repositories | ✓ | ✓ | ✓ | ✓ |
| Fleet dashboard | ✓ | ✓ | ✓ | ✓ |
| Canary releases & rollback | ✓ | ✓ | ✓ | ✓ |
| Immutable tags | ✓ | ✓ | ✓ | ✓ |
| Support | Community | Community + email | Priority email | SLA |
Pulls beyond the included bandwidth are billed at $0.09/GB. Storage is capped at the included amount, and upgrading raises it. The trial stops at its limit rather than billing you.
The NexusRun CLI and runtime are Apache 2.0 and free forever. Cloud is optional: you can always push to any OCI registry you already run.
Yes. The nexus CLI and runtime are Apache 2.0 and stay free, with no local host limits. Cloud plans cover private artifact delivery and Fleet management, with enrolled-host, storage, and bandwidth limits. Local inference has no per-token bill.
Yes. Open your organization and select Fleet to see enrolled hosts and their reported versions and test results. Owners and administrators can start canary releases, pause rollouts, request tested rollbacks, and revoke hosts. Members can view status. Each host needs its local runtime, enrollment, and update watcher.
Your live inference inputs, generated answers, and secret values stay local. Cloud stores artifacts you publish, including their agent instructions and test fixtures. Sealed units, required for Fleet, include model weights. Fleet also records assignments, hardware totals, last check-ins, and acceptance counts or failure codes. Ordinary linked units let you distribute weights separately.
The registry recomputes the SHA-256 of every blob it receives and rejects anything that doesn't match the digest you claimed. Access tokens are stored only as hashes, and the short-lived tokens that authorize each pull or push expire in five minutes.
An active Fleet release keeps running while disconnected. New assignments and reports require a connection to Cloud. For permanently air-gapped machines, use an approved physical transfer: the offline bundle guide includes NexusRun and a pinned llama.cpp runtime for a compatible Linux host.
No Docker or Kubernetes runtime is required. Fleet hosts need NexusRun, the release's exact managed CPU runtime, and an outbound update watcher. Inference runs locally; Cloud coordinates which release each host should test and activate.
Today: Linux on x86-64, CPU inference. nexus doctor tells you what a given machine can actually use, not just what it detects.
Yes, any time. nexus push works with any OCI registry (GHCR, Harbor, ECR, and others), and anything you store with us can be pulled back out with standard tools.
Both package agents as OCI artifacts. cagent needs Docker on the target. NexusRun doesn't, which is what matters on locked-down, embedded, or air-gapped machines.
Create an account, prepare and enroll your hosts, then manage releases from the Fleet tab in your organization.
Fleet or air-gap rollout? Email hello@nexusrun.dev for a 20-minute demo.