Private beta now for ROS 2 remote debugging teams.
RoboTunnel is currently in invite-only beta for teams that need ROS 2 remote debugging that actually works. Invited beta users become Founding Developers and receive a lifetime discount when paid plans launch. The current packaging direction is active robots plus relay-heavy connection usage, not seats, messages, prompts, or LLM usage.
Private beta
Approved users are invited manually. The goal of this phase is to validate one sharp workflow: managed ROS 2 debugging, with Debug Projection and first-line diagnosis on difficult field networks.
- Invite-only access via direct beta request
- Free during the beta period for invited users
- Founding Developer status for invited beta users
- Lifetime discount when paid plans launch
- Best fit: field robots, ROS 2, NAT, unstable onsite networking
- Current launch promise: install, start a managed debug session, and run first diagnostics
- Direct feedback loop with the product team
Packaging candidate after beta
Final pricing is not published yet, but the current plan is intentionally narrower and more concrete than generic connectivity pricing. The likely paid surface is active robots plus relay-heavy connection usage required to keep those debug workflows dependable.
- Likely pricing axis: active robots
- Likely overage axis: relay-heavy usage
- No LLM API markup planned
- No seat-based pricing as the primary model
- No per-message or prompt billing model planned
- Agent remains open source permanently
- Detailed tiers will be announced after beta
What does v0.3.0 formally cover?
Linux robot hosts, ROS 2 remote debugging, one documented install path, first-session connectivity, Debug Projection basics, and a clear troubleshooting flow when direct paths fail.
Who is the best fit?
Small robotics teams with real field robots, ROS 2, bad remote networks, and frequent remote-debug needs. If you mainly need generic IoT fleet management, this launch is probably too robotics-specific.
What stays open source?
The robot-side agent remains open source. The hosted platform, signaling, auth, routing, and operational control plane are managed services, not part of the open-source release.
How do LLM costs and self-hosting work?
You bring your own LLM key and the key stays on the robot. Advanced platform overrides exist for internal setups, but supported v0.3.0 onboarding is managed platform plus open-source agent.
Questions about rollout or access?
The beta is intentionally high-touch right now. Reach out with your robot platform, ROS distribution, network situation, and current remote-debug pain point.
Request beta access