# DeepSeek Harness Update Enhances Multimodal Capabilities

- Category: Flash (快讯)
- Published: 2026-08-20T12:00:00.000Z
- Source date: 2026-08-20
- HTML: https://www.88z.ai/en/news/flash-20260820-27c442aae22b6a2b
- Markdown: https://www.88z.ai/en/news/flash-20260820-27c442aae22b6a2b.md
- Other language: https://www.88z.ai/news/flash-20260820-27c442aae22b6a2b.md

DeepSeek Harness releases version v0.1.0-rc.8, focusing on enhanced multimodal support with native image requests and mixed text-image inputs, while allowing Claude Code and Codex to be called as sub-agents. The update also improves tool concurrency and Windows terminal experience, fixes stability issues, and reinforces its role as a unified agent scheduling layer.

## At a glance

- Supports native image requests and mixed text-image inputs, with ability to call Claude Code and Codex as sub-agents
- Optimizes tool concurrency and Windows terminal experience, fixing stability issues to reinforce its scheduling layer position

## The update

DeepSeek Harness has officially launched version v0.1.0-rc.8, with this update focusing on significantly enhancing multimodal processing capabilities. The new version not only supports native image request functionality but also enables seamless handling of mixed text and image inputs, providing richer interaction methods for complex tasks. In terms of agent integration, users can now flexibly call Claude Code and Codex as sub-agents based on actual needs, further expanding the system's capability boundaries. Beyond functional upgrades, this version includes several optimizations in performance and user experience. By improving the tool concurrency query mechanism, processing efficiency has been boosted. Additionally, the user experience on the Windows terminal has been specifically optimized, and multiple previous stability issues have been resolved. These improvements further solidify DeepSeek Harness's technical position as a unified scheduling layer for agents, making it more stable and efficient when coordinating and managing various AI agents.

## Sources

- [机器之心](https://mp.weixin.qq.com/s/D-kkE__o9yt8QVzQLUlvOw) (2026-08-20)

This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.