<div align="center">
<p>
<a href="https://trendshift.io/repositories/25042?utm_source=repository-badge&utm_medium=badge&utm_campaign=badge-repository-25042" target="_blank" rel="noopener noreferrer"><img src="https://trendshift.io/api/badge/repositories/25042" alt="1jehuang/jcode | Trendshift" width="250" height="55"></a>
</p>
<picture>
<source media="(prefers-color-scheme: dark)" srcset="assets/brand/jcode-mark-inverse.svg">
<img src="assets/brand/jcode-mark.svg" alt="jcode halftone donut logo" width="160">
</picture>
# jcode
<p>
<strong>High-performance coding agents.</strong><br>
Run dozens of agents in parallel without melting your machine. More agents, less RAM, no lag.
</p>
[](https://github.com/1jehuang/jcode/commits/master)
[](https://github.com/1jehuang/jcode/stargazers)
[](https://github.com/1jehuang/jcode/releases)
[](https://discord.gg/nBe9vGyK9a)
[](https://www.ycombinator.com/companies/jcode)
[](https://x.com/1jehuang)
</div>
<div align="center">
<a href="https://github.com/1jehuang/jcode/releases/download/readme-assets/jcode-performance-demo-2026-10.mp4">
<img src="https://github.com/1jehuang/jcode/releases/download/readme-assets/jcode-performance-demo-2026-10-30fps.webp" alt="jcode performance demonstration" width="900">
</a>
<p><em>jcode performance demonstration (1x speed)</em></p>
<a href="https://github.com/1jehuang/jcode/releases/download/readme-assets/jcode-yc-launch.mp4">
<img src="https://github.com/1jehuang/jcode/releases/download/readme-assets/jcode-yc-launch.webp" alt="jcode YC launch video" width="800">
</a>
</div>
<div align="center">
## Installation
</div>
```bash
# macOS & Linux
curl -fsSL https://jcode.sh/install | bash
```
<div align="center">
## Benches
</div>
### RAM per additional session (headless)
| Tool | First session | Each additional session | Comparison (additional) |
|---|---|---|---|
| **jcode** | **71.3 MB** | **~3.3 MB** | baseline |
| **OpenCode** | 637.1 MB | ~5.9 MB | **1.8× more RAM** |
| **pi** | 84.0 MB | ~79.6 MB | **24.1× more RAM** |
| **Claude Code** | 272.9 MB | ~169.3 MB | **51.3× more RAM** |
| **Codex CLI** | 398.6 MB | ~194.1 MB | **58.8× more RAM** |
[Reproduce](scripts/bench_headless_memory.py).
### Time to first input
Time until typed text appears on screen, across 10 interactive launches.
| Tool | Time to first input | Comparison |
|---|---|---|
| **jcode** | **19.1 ms** | baseline |
| **pi** | 596.4 ms | **31.2× slower** |
| **Codex CLI** | 905.8 ms | **47.4× slower** |
| **OpenCode** | 1047.9 ms | **54.9× slower** |
| **Claude Code** | 3512.8 ms | **183.9× slower** |
Full numbers on [jcode.sh](https://jcode.sh).
<div align="center">
## Screenshots
[Browse the interactive gallery on jcode.sh](https://jcode.sh/#screenshots).
<table>
<tr>
<td align="center"><a href="https://jcode.sh/screenshots/inline-image-rendering.webp"><img src="https://jcode.sh/screenshots/thumbs/inline-image-rendering.webp" alt="Jcode rendering screenshots inline after reading image files" width="440" /></a><br>Inline images</td>
<td align="center"><a href="https://jcode.sh/screenshots/resume-menu.webp"><img src="https://jcode.sh/screenshots/thumbs/resume-menu.webp" alt="Jcode resume menu with a session list and preview" width="440" /></a><br>Resume menu</td>
</tr>
<tr>
<td align="center"><a href="https://jcode.sh/screenshots/latex-math-rendering.webp"><img src="https://jcode.sh/screenshots/thumbs/latex-math-rendering.webp" alt="Jcode rendering LaTeX mathematics inline" width="440" /></a><br>Inline LaTeX</td>
<td align="center"><a href="https://jcode.sh/screenshots/cache-cold-notification.webp"><img src="https://jcode.sh/screenshots/thumbs/cache-cold-notification.webp" alt="Jcode warning that the prompt cache went cold" width="440" /></a><br>Cache cold notification</td>
</tr>
</table>
</div>
## Swarm
Spawn two or more agents in the same repo, and they will automatically be managed by the server to allow native collaboration. When agent A edits a file that agent B has read (code shifting under its feet), the server notifies agent B. Agent B can ignore it if it is not relevant, or it can check the diff to make sure that it doesn't conflict. Each agent has messaging abilities, capable of DMing just one agent, broadcasting to all other agents hosted by the server, or just agents working in that repo. This allows you to spawn multiple sessions in the same repo, and have all conflicts automatically resolved.
Agents are also able to spawn their own swarms autonomously. They have a swarm tool which allows them to spawn in their own teamates to accomplish tasks in parallel. Doing so turns the main agent into a coordinator and the spawned agents into workers. Groups of agents, their messaging channels, their completion statuses, etc are all automatically managed. This can be done headlessly or headed.
## OAuth and Providers
Supports a [wide range of providers](https://jcode.sh/docs#providers), including Claude OAuth and OpenAI Codex OAuth.
## Customizability / Self-Dev
Jcode is inventing a new form of customizability. One that doesn't limit you to what a plugin or extension can do. Tell your jcode agent to enter self dev mode, and it will start modifying its own source code. Jcode is optimized to iterate on itself. There is significant infrastructure around self development, which allows it to edit, build, and test its own source code, then reload its own binary and continue work in your (potentially many) sessions, fully automatically.
<!-- Add self-dev demo thumbnail/video and fuller writeup here. -->
## Other
- [Token optimization](docs/HARNESS.md#token-optimization)
- [KV cache](docs/HARNESS.md#kv-cache)
- [Hill-climbable measurements](docs/HARNESS.md#hill-climbable-measurements)