Visit website ↗
DeepSeek’s open-source agent client for working with files, running tools and composing custom workflows.
TOOLORA / Local reading edition
DeepSeek Harness
DeepSeek’s open-source agent client for working with files, running tools and composing custom workflows.
Overview
DeepSeek Harness connects models to a local workspace and execution tools for repository inspection, code changes, commands and document tasks. It offers a desktop app and a self-hosted web interface. Developers can compose models, tools and interfaces through Cordis plugins. This is an agent runtime and client; a model service must still be configured or accessed through an account.
- Company country / region
- China · Company registration
Access options
Visit website ↗
DeepSeek’s open-source agent client for working with files, running tools and composing custom workflows.
Usage guide & costs
- Before you start
Public preview; features and interfaces may change. Review file permissions and commands before execution. Official models follow platform billing; custom-model costs and data processing depend on the selected provider. A locally running interface does not imply offline inference.
Official details ↗ · Official details ↗
Related product news
News summaries are prepared by this site; open the title to read the original.
DeepSeek updates V4.1-Flash-1017
DeepSeek upgrades its web, app and API to V4.1-Flash-1017, retaining the deepseek-flash model ID and pricing. The update targets reasoning and agent tasks, with optimizations for DeepSeek Harness, Tencent CodeBuddy and WorkBuddy.
DeepSeek ↗Product profile →Company news
DeepSeek opens its in-house search API
DeepSeek opens Search / Fetch APIs for its own search engine, providing multilingual web search and page retrieval for models and agents.
Products covered · DeepSeekDeepSeek ↗Product profile →
DeepSeek releases V4.1-Flash
DeepSeek releases V4.1-Flash with native visual understanding and an asymmetric MoE architecture aimed at reducing inference and cache costs. The multimodal API uses deepseek-flash; older Flash model IDs temporarily route to the new model.
Products covered · DeepSeekDeepSeek ↗Product profile →
Official sources and references
- Official websitehttps://www.deepseek.com/harness/
- Official websitehttps://github.com/deepseek-ai/deepseek-harness
- Documentationhttps://deepseek-harness.github.io/deepseek-harness/
- Pricing sourcehttps://api-docs.deepseek.com/quick_start/pricing
Descriptions summarize public information; consult official websites for features, pricing and regional availability.