Jcode

Jcode provides high-performance coding agents written in Rust, designed to run many parallel AI-assisted development sessions without degrading local machine performance. The tool launches sessions in under 50 milliseconds and uses roughly 48× less RAM per headless session than Claude Code, while supporting major AI providers including Claude, OpenAI, Gemini, and local servers. It offers a TUI, non-interactive runs, persistent background servers, and resumable sessions for demanding multi-agent workflows.

Traction –/10

HQ unknown

Developer Tools
AI Agents
Artificial Intelligence
Software Only

Updated 9 days ago

Funding

Raised to date

$500KRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.

Funding rounds are not available yet.

Founders

Founder details are not available yet.

Product

Problem

Heavy development workflows are bottlenecked by the harness's performance and resource efficiency. Spawning too many agents can cause a machine to throttle or crash, and a slow, resource-hungry harness keeps developers from leaning into parallel work across multiple sessions.

Solution

Jcode delivers high-performance coding agents written in Rust, built to stay out of the machine's way. An extra session costs megabytes instead of hundreds of megabytes, and a new session is ready for input in under 50 milliseconds, so developers can run many agents at once on hardware they already own. The tool installs with a one-line command, connects to models developers already pay for (including Claude, OpenAI, Gemini, Copilot, and local servers), and supports a TUI, non-interactive runs, persistent background servers, and resumable sessions. Configuration supports named provider profiles, project-level instruction files, MCP servers, and a skills system that injects markdown instruction packs based on conversation matching.

Target Audience

Primary users are software developers and engineering teams running demanding, multi-agent development workflows who need to maximize parallel agent execution on local hardware without performance degradation.

Features

Written in Rust; benchmarks show 48× less RAM per additional headless session (3.5 MB vs 168.6 MB for Claude Code) and 20× less in interactive mode (10.4 MB vs 212.7 MB)
Time to first input of 48.7 ms, compared to 3,512.8 ms for Claude Code (72× faster)
Spawn-mode and concurrency-cap configuration for running "swarm" agent fleets in parallel
Built-in login flows for Claude subscription, OpenAI, Gemini, GitHub Copilot, Azure OpenAI, Ollama, LM Studio, plus any OpenAI-compatible endpoint via named profiles
MCP server support with on-disk schema caching so slow handshakes never block startup or invalidate prompt cache
Skills system with vector-based conversation matching for context-aware injection of instruction packs
Headless and remote login flows, including --no-browser auth URL printing for automation
Configuration via TOML with sections for provider, display, features, keybindings, agents, hooks, notifications, and ambient background mode