Comparison
Sesori vs Claude Code Remote Control
Claude Code Remote Control is Anthropic's built-in way to stream a local Claude Code session to your phone or browser. Sesori is a separate mobile cockpit that today targets OpenCode and is designed to support more agents over time. The two solve overlapping problems with different tradeoffs around vendor lock-in, pricing tiers, and privacy posture.
TL;DR
If you live entirely inside Claude Code on an eligible Anthropic plan and want the most official path, use Remote Control. If you use OpenCode (or want a mobile client that does not depend on an Anthropic subscription), Sesori is the better fit.
Side by side
| Feature | Sesori | Claude Code Remote Control |
|---|---|---|
| Agents supported | OpenCode today. Claude Code, Codex, and others on the roadmap. | Claude Code only. |
| Required AI plan | None. Bring your own OpenCode setup and model keys. | Eligible Anthropic plan: Pro, Max, Team, or Enterprise. API keys are not supported. Team and Enterprise need admin enablement. |
| Bridge / pairing model | Local open-source bridge CLI on your laptop, paired to the app via account login. | Built into the Claude Code CLI. Pair via QR code from the desktop terminal. |
| Relay visibility | End-to-end encrypted relay. Sesori only sees opaque encrypted traffic, not your code, prompts, or responses. | Traffic routes through Anthropic infrastructure. Anthropic terms and policies apply. |
| Native mobile app | Native iOS and Android app focused on AI coding workflows. | Uses the Claude consumer iOS/Android app or any browser. Chat-centric interface. |
| Session lifecycle | Bridge runs while you want phone access. Reconnects when your phone hops networks. | The terminal that started the session must stay open. Close the terminal and the session dies. |
| Parallel sessions | Designed to grow into multi-agent and multi-session workflows. | One remote session per interactive Claude Code process. Server mode supports concurrent sessions up to a configurable capacity. |
| Voice coding | First-class voice input designed for hands-free coding sessions. | Voice via the Claude consumer app, not specialised for coding workflows. |
| Maturity | Early access, narrow scope (OpenCode-first), iterating quickly. | Shipped as part of the official Claude Code product. |
Which one is right for you?
Pick Sesori if you
- Use OpenCode, or expect to use multiple AI coding tools over time.
- Want your mobile client to not depend on a specific vendor plan.
- Want an open-source bridge you can read, audit, and self-host.
- Care about a relay that cannot see your code, prompts, or responses.
- Want a mobile app designed specifically for coding, not a chat sidebar.
Pick Claude Code Remote Control if you
- You exclusively use Claude Code and are on an eligible Anthropic plan.
- You want the most official path, sanctioned by Anthropic's product team.
- You are comfortable with Anthropic relaying your session traffic.
- You prefer the Claude consumer app's chat-style mobile UX.
FAQs
No. Sesori is an independent project. It started with OpenCode support and the team plans to add Claude Code, Codex, and other agents through a plugin architecture. Claude Code Remote Control is a feature inside the official Claude Code product, owned by Anthropic.
No. Sesori is the mobile client for whichever AI coding tool you already use. Today that is OpenCode, which can talk to your own model providers. Sesori itself does not gate access behind a Claude plan.
No. Traffic between the bridge on your laptop and the Sesori mobile app is end-to-end encrypted with X25519 and XChaCha20-Poly1305. The relay forwards opaque binary. There is no plaintext access to code, prompts, or AI responses.
That is on the roadmap. The bridge is being designed so that additional agent integrations can be added over time, including Claude Code, Codex, and others. OpenCode is the first supported agent because the team wanted to ship one polished integration before fanning out.
Both keep your code on your laptop. The main difference is the relay. Sesori's relay only sees encrypted traffic by design. Claude Code Remote Control routes through Anthropic, so the trust model is "trust Anthropic to handle session relay correctly under their terms".