AgenticSeek at a glance
Checked 16 Oct 2026AgenticSeek (27.5k stars, GPL-3.0, very active in October 2026) picks a browsing, coding, file or planning agent for each request, searches through a bundled SearxNG, fills web forms and runs code in Python, C, Go, Java and more. It runs in Docker with a web UI, or as a CLI. It is designed for local models (at least a GPU that can run a 14B model); cloud APIs such as OpenAI or DeepSeek are optional. Voice is listed as in progress. Largely a one-maintainer side project with no roadmap or funding. No slides, site builder, scheduler or app connectors.
RepoQuarry. “AgenticSeek.” Checked 16 Oct 2026. https://repoquarry.dev/projects/agenticseekHow much of Manus does it replace?
2 of 6 jobs. Local browsing, coding and file tasks; no slides, websites, schedules or app connectors.
For deep research reports, slides generated from a prompt, scheduled agent tasks, connectors to your apps, see other Manus alternatives. You may also need to build an interface, an API server and hosting yourself.
Facts, with sources
Every answer links to where we found it. We read the documentation; we did not install it.
Everything runs on your machine — no cloud, no data sharing.github.com · agenticSeek#readme ↗ today
provider_name = ollama # or lm-studio, openai, etc..github.com · agenticSeek#readme ↗ today
API Key are totally optional for user who choose to run LLM locally.github.com · agenticSeek#readme ↗ today
At a minimum, a GPU capable of running Magistral, Qwen or Deepseek 14B is required.github.com · agenticSeek#readme ↗ today
./install.bat # windowsgithub.com · agenticSeek#readme ↗ today
This step will download and load all Docker images, which may take up to 30 minutes.github.com · agenticSeek#readme ↗ today