DeepSeek releases official desktop apps for Harness v0.2 preview
DeepSeek has released official desktop applications for its open-source agent harness, DeepSeek Harness (dsh), accompanying the v0.2 preview. The installers are available for macOS (Apple silicon) and Windows (64-bit) systems, and can be downloaded from deepseek.com/harness or run via npx @deepseek-ai/dsh web. The release is marked as a developer preview, with DeepSeek warning that…
Key points
- DeepSeek released official desktop apps for Harness v0.2 preview on macOS and Windows
- The v0.2 preview includes plugin management, document work, automation tasks, and file review
- DeepSeek-V4-Flash-0731 scored 82.7 on Terminal Bench 2.1 using dsh in minimal mode
The v0.2 preview includes built-in capabilities for office and development tasks, plugin management, file and diff review, document work features, and an automation task plugin for scheduling recurring prompts. The desktop build bundles the dsh command, eliminating the need for separate Node.js or pnpm installations. Users can sign in with a DeepSeek account or add an API key, with account models gaining web search without an extra key per rc.1 notes.
Architecturally, DeepSeek emphasizes that everything in dsh is a plugin, including the model adapter, tool registry, and agent loop, allowing any layer to be replaced. The harness is not locked to DeepSeek models and supports third-party providers via OpenAI-compatible endpoints. The repository has garnered 240,000 GitHub stars and 29,000 forks. Additionally, dsh serves as DeepSeek’s evaluation rig, where DeepSeek-V4-Flash-0731 scored 82.7 on Terminal Bench 2.1 in minimal mode. The v0.2.1-alpha.1 build adds an experimental Claude Code Mods compatibility layer, a ‘Let Agent create a plugin’ entry, a Developer Tools bundle, and a --public-url flag, while removing runtime invariant plugins as a breaking change.
The story so far
2 episodes →- DeepSeek releases official desktop apps for Harness v0.2 previewthis story
DeepSeek Harness v0.2 Brings Official Desktop Apps to Its Open-Source Agent Harness
MarkTechPost · 4 October 2026
Loading the full article…
This text was published by MarkTechPost and written by Michal Sutter. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Agents & Tools
All →- Near AI Cloud enables private GitHub Copilot inference through VSCode extension · 1 src
- NVIDIA IsaacTeleop tutorial shows how to build a retargeting engine in NumPy · 1 src
- Cloudflare launches competition to build agent-focused Git platform · 1 src
- Show HN: Pi pod lets users Run Pi coding agent in self‑hosted sandboxes · 1 src
- Instinct hits $10B valuation as text-based AI agents expand · 1 src
Comments
via GitHub Discussions