Community waitlist · Early access
We are building a one-click desktop app for the open-source DeepSeek Harness — local execution, your own API key, no terminal setup. The app is not available yet. Join the waitlist to reserve early access.
Local by design — code and credentials stay on your machine; the app plans to use your own model key.
This is an unofficial community waitlist page. We are not affiliated with DeepSeek, and nothing here is a commitment that the app will ship. Joining the waitlist costs nothing.
What it is
It is an open-source agent harness from DeepSeek AI, released as a v0.1 developer preview under the MIT license. A harness is the runtime that turns a language model into a working agent: it gives the model tools such as file editing, shell commands and search, manages sessions, applies permission policies, and records every step so a run can be inspected, resumed or replayed.
Its design principle is that everything is a plugin: models, tools, skills, sessions, sandboxes, storage, loops, scheduling and the UI are all plugins on the Cordis framework. It ships four presets — Standard, Code, Minimal and Creator — and writes an append-only session log that powers a trajectory view where you can inspect prompts, reasoning, tool calls and results by source.
The official way to run it today is a local command, npx @deepseek-ai/dsh web, which assumes Node.js, a working terminal and a configured model provider. The Desktop app aims to handle the installation, setup and update steps, so the open-source engine stays the same underneath.
Pricing hypothesis
Prices shown here are a working hypothesis, not an offer. Nothing is being sold yet.
Early access
We are using this waitlist to understand demand before building. Qualified testers may later receive an invitation to a secure, refundable reservation or a paid beta — no payments are collected on this page today.
After joining, we may ask a few short questions about your setup and your pain points. Your answers help us decide what to build first.
FAQ
A coding harness is the runtime around a language model that makes it a working agent: tool execution (file editing, shell, search), session management, permission policies and a traceable log. DeepSeek Harness is an open-source harness where every one of those capabilities is a replaceable plugin.
It is a community waitlist for the Desktop app we plan to build around the open-source project. The app is not available yet. Joining the waitlist reserves early access, costs nothing, and tells us whether there is real demand before we invest further.
The source code is free under the MIT license. Running an agent still needs model access: you configure a provider such as the DeepSeek API with your own key, and tokens are billed by that provider. The desktop app does not change that — you keep using your own key.
Yes. The official documentation covers other providers and custom OpenAI-compatible endpoints. Because the model adapter is itself a plugin, swapping models does not require changing the framework source.
Not yet. DeepSeek labels it a developer preview and explicitly warns that compatibility-breaking changes are coming. The project changes quickly, so anything built on top of it needs to track those changes deliberately. The waitlist gives us a way to talk to early users before we commit to a release date.