CLUSTER DESKTOP-AGENTS-LOCAL · 32 ENTRIES · EARLIEST ENTRY 2026-08-25

Local AI model workbenches

This cluster comprises desktop applications for running, testing, and tuning local AI models, often with agent execution and user-selected model providers. The dated history is too thin to read a shape from, with only four dated entries before the index started, so the cluster predates the index or its members are largely undated. On traction, dsh-desktop leads with 6,823 stars, followed by vibe-holding at 201 and laobos-agent-studio at 12.

Entry timeline

ENTRY TIMELINE

NOV 2025: 1APR 2026: 1AUG 2026: 6SEP 2026: 16OCT 2026: 8
NOV 2025OCT 2026

Members by activity

Activity signals, not revenue.

Build origin · 1 hand-coded · 19 unlabelled

Operator · 11 hobby · 1 indie · 1 commercial · 1 open source · 6 unlabelled

Two labels per entry, never collapsed into one: how it was built, and who stands behind it. Counted over the 20 listed below. Unlabelled means nothing has decided yet — not that the answer is small.

  1. 01dsh-desktop6.8k★25 daysliveDesktop application that launches a local AI agent environment, manages model provider settings, and preserves sessions for building agent workflows.
  2. 02laobos-agent-studio12★28 days—Desktop client that lets users interact with AI agents locally.
  3. 03beichen-pi-desktop16★1 mo—indieLocal-first desktop agent platform that gives users explicit control over AI model context lifecycles.
  4. 04Belmont-agent-harness2★1 moliveDesktop app that runs AI agents on local files with multiple model providers and approval-gated access to tools.
  5. 05--agent2★10 dayslivehobbyDesktop AI agent client that separates confirmed memories from inferred impressions and streams tool calls from a local runtime.
  6. 06zimage-studio1★7 days—hobbyDesktop workbench that installs, launches, and monitors a local AI image generation environment, including queue and VRAM controls and model fine-tuning.
  7. 07Mework1★7 dayslivehobbyLocal-first desktop workbench for AI agents that executes tools on the user's machine and connects to user-chosen model providers.
  8. 08LocalAgent-UI1★2 dayslivehobbyIn-browser AI agent workbench that runs models locally, visualizes reasoning and tool calls, and executes code in an isolated sandbox.
  9. 09Hy-MT2-Windows1★21 dayslivehand-codedhobbyDesktop app that runs a local AI translation model on NVIDIA graphics cards and offers batch translation and an API for other applications.
  10. 10remuda5★1 mo—Desktop application that lets users run, test, and tune local AI models.
  11. 11Maxolla1★17 days—Desktop application for managing local AI models on Windows, offering model installation, chat testing, GPU monitoring, and backup restore.
  12. 12Monolith6★1 mo—hobbyDesktop AI client that connects to multiple providers with personal API keys, runs local models, and manages cooperating AI agents.
  13. 13molly-design0★12 days—hobbyLocal desktop graphic design workbench with an AI agent that creates and revises posters on an editable canvas and exports images.
  14. 14mxstage0★9 dayslivecommercialLocal AI workbench that lets users review, approve, and commit corrections to IBM Maximo data with human approval before changes are written.
  15. 15hangar0★today—hobbyDesktop utility that manages local AI model servers on a Mac, providing a single web page, command-line interface, and OpenAI-compatible endpoint to start, stop, and monitor them.
  16. 16wbBridge0★9 days—hobbyTray utility that supplies a third-party AI assistant with free models by exposing a local API and syncing model configurations.
  17. 17ollama-gui0★13 dayslivehobbyDesktop app that manages local AI models by browsing, pulling, updating, inspecting, customizing, and testing them without a terminal.
  18. 18python-aiconfigurator0★21 daysliveopen sourceTool that plans and generates configuration files for running AI models offline.
  19. 19EasyAgent0★1 day—Desktop application that runs AI agents locally using the user's own models and machine.
  20. 20test-your-astra0★9 days—hobbyLocal desktop tool that probes AI models with parallel test prompts and automated visual scoring to detect performance degradation.