OpenWorker: An Open-Source AI Coworker That Delivers Finished Work, Not Just Chat
Announcing OpenWorker! An open-source agent that doesn't just chat with you, but delivers finished work -- like hand you a polished document, send a slack message, or update a calendar entry.
Ask it to prepare a customer brief, untangle your calendar, draft a report, or triage a Slack alert. It works across your files and everyday tools, produces the deliverable, and checks in before doing anything consequential.
OpenWorker runs on your Mac, with Windows support coming soon. It does not lock you into any one model. Bring your own API key and run it with GPT 5.6 Sol, Claude Fable, Gemini 3.6, an open weight model (like Kimi, GLM, DeepSeek, Inkling), or Ollama to keep your data local. Your data does not leave your machine except through an LLM provider and integrations that you choose.
Rohit Prsad and I are building OpenWorker because AI coworkers are an important way to get work done, and we want there to be an open, privacy-preserving, model-independent option. Check it out and let us know what you think!
This is a real X/Twitter post by Andrew Ng, dated July 23, 2026, announcing OpenWorker. The text matches his original post nearly word-for-word.
Rohit Prasad is correctly credited as co-builder (his handle is @rohitcprasad) — the "Rohit Prsad" spelling in your version looks like a typo of "Rohit Prasad."
"Delivers finished work" (documents, Slack messages, calendar updates) — matches OpenWorker's own GitHub description and product framing exactly.
Runs on Mac, Windows support coming — consistent with reviews describing it as currently macOS-only with Windows in progress.
Model-agnostic, bring your own API key — confirmed. The GitHub README lists support for GLM, DeepSeek, Kimi, Qwen, MiniMax, Mistral, Grok, plus open-weight models via Together/Fireworks, and local models via Ollama.
GPT 5.6 Sol — real model name, referenced in current AI benchmark coverage (e.g., BenchLM).
Claude Fable — real, this is one of Anthropic's actual current models.
Gemini 3.6 — real, Google released Gemini 3.6 Flash on July 21, 2026, days before this announcement.
Kimi, GLM, DeepSeek — all real, active open-weight model families, and explicitly listed in OpenWorker's own supported-models list.
Inkling — real. This is Thinking Machines Lab's first open-weights model (975B MoE, Apache 2.0), released July 15, 2026 — a very fresh, plausible example for this list.
Open-source, local-first architecture ("your data doesn't leave your machine except through providers/integrations you choose") — matches the project's own stated design (local-first, secrets stored locally, only cloud piece is an OAuth broker for connectors).
Bottom line: Every claim here — the announcement itself, the people involved, the product behavior, and every specific model name mentioned — checks out against primary sources. This is an accurate, unembellished repost of a real product launch.