Bedcoder vs CoreAI Zoo

Bedcoder is a phone-side companion for vibe-coding. When your Claude Code is running long tasks on your laptop — installing dependencies, running test suites, waiting on a build, scaffolding a new feature - you don't have to sit at the desk to keep tabs on it. Pair Bedcoder once via QR code, and you can drive the same agent session from your phone. What you can do from the phone: • Chat with the agent in real time. Send new prompts, see streaming responses. • Review and respond to the agent's proposed actions (e.g. approve or skip a tool call). • Mirror the terminal. Watch live output from commands running on your laptop. • Preview running dev servers in a remote browser, with secure tunnels back to your computer. • Browse and read project files inside your active project directory. • View your laptop screen when the agent is using a GUI app. How it works: The Bedcoder daemon runs next to your existing AI coding CLI on your computer (Currently Claude Code is supported). The phone app pairs with it via a single QR scan. Every message between the phone and the laptop is end-to-end encrypted with XSalsa20-Poly1305 — the relay server never sees your prompts, your code, your responses, or your encryption key. Session interop: Bedcoder shares session storage with your desktop CLI. A conversation started at your desk picks up cleanly on your phone, and vice versa — no copy-paste, no context loss. Privacy by design: • End-to-end encrypted between your phone and your laptop daemon. • Zero-knowledge relay: no accounts, no message history, no telemetry. • Filesystem access is sandboxed to your active project directory. Bedcoder is open source. Requires the Bedcoder desktop daemon running alongside your AI coding CLI (Claude Code is supported now. Codex is our next goal).

CoreAI Zoo puts the open-weight AI model zoo on your iPhone, iPad and Mac. Pick a model, download it once, and run it entirely on your device. No account, no server, and nothing you type, say or photograph ever leaves the device. CHAT Stream answers from open language models such as Qwen3, Gemma, LFM2.5, Granite and Mistral, from 0.6B to 30B+ parameters depending on your device's memory. Vision-language models (Qwen3-VL, MiniCPM-V, LFM2.5-VL) answer questions about your photos. VISION Ask questions about an image, read text with on-device OCR, and describe what the camera sees. AUDIO Transcribe speech with Whisper large-v3-turbo, Parakeet and Qwen3-ASR. Talk to an audio-understanding model, generate speech with Kokoro and VoxCPM, and try on-device music generation. CAMERA Live depth estimation, object detection (YOLOX, RF-DETR), action recognition and super-resolution on the camera feed. TOOLS Experimental models such as diffusion-based language models. BENCH Measure load time, tokens per second and memory for any chat model on your own hardware. HOW IT WORKS Every model runs through Core ML on Apple silicon. Weights are downloaded from the Hugging Face Hub on first use over HTTPS, then cached, so the app works offline afterwards. One model is resident at a time to stay within memory limits. New models arrive through a live catalog without an app update. GETTING STARTED Start with Qwen3-0.6B (454 MB) for the fastest first run. Larger models state the free space and memory they need; a recent iPhone Pro or an Apple silicon Mac runs the mid-size ones comfortably. Model weights follow their own licenses, listed on each model's Hugging Face page. The app itself is BSD-3-Clause and part of the open CoreAI model zoo. Report issues and request models at github.com/john-rocky/coreai-model-zoo.