Factory

Droids can now pursue goals autonomously over multi-day horizons. You describe what you want, approve the plan, and come back to finished work. We call these Missions.

Read more here: docs.factory.ai/enterprise/net…

Factory was built from day one to deploy wherever your code already lives. On laptops, in CI pipelines, on VMs, inside Kubernetes clusters, and in networks with zero outbound internet connectivity. We support three deployment patterns, and you can mix them across teams: 1/ Cloud-managed 2/ Hybrid 3/ Fully air-gapped @nvidia, the US Government, and the world's largest financial institutions run Factory self-hosted today.

I'm becoming a real Droid fan now! Here we have Calude Code vs Droid in Lego San Andreas glm-5-turbo edition! - 0-shot - same prompt, same model, same reasoning - stealing vehicle doesn't work in cc version - time of the day simulated in Droid version! Look at the image! 🤯

I have never used in deep Droid, but I'm really impressed by just first experiments 🤯 Are some heavy users out there that used other coding harnesses too? Is it really so much better? I will start using it more and more!

Droid is really powerful 👀 Where did I live so far? @FactoryAI what's the magic behind this beast?

GPT-5.4 Mini has an impressive cost-to-capability ratio and punches above its weight on practical multi-step engineering work. Other strengths include:    •  Strong step-by-step planning    •  Cleaner, more reliable tool use    •  Better persistence through debugging and blockers    •  Clear gains from higher reasoning effort

GPT-5.4 Mini is now available in Droid.

Come meet the team behind the Droids at @Nvidia GTC. Booth #4053 on the 2nd floor in the ballroom

Read more: factory.ai/news/factory-a…

We built Factory Analytics alongside the frontier enterprise teams deploying Droids across large engineering organizations. When thousands of engineers start using agents, leadership needs visibility into adoption, cost, and output. Analytics is the first step towards a better control plane for agent-native development.

Today we’re launching Factory Analytics. Enterprise teams can now see exactly how AI agents translate into engineering outcomes: tokens → usage → commits → pull requests → shipped software. The missing layer for proving ROI in agent-native software development.

Free 200M Tokens. Let's Build Something With Factory AI. Live From SF x.com/i/broadcasts/1…

To get started, install the Figma MCP server: / mcp + Add MCP server from registry -> Add Figma Then prompt Droid to send a page from your local web app to Figma. See the docs for more: #quick-start-add-from-registry" target="_blank" rel="nofollow noopener">docs.factory.ai/cli/configurat…

Factory can now push what you’re building directly from code into @Figma. Take a working prototype from the codebase, send it to a Figma canvas, and let designers and PMs edit components directly.

the new “mission” (preview) feature in @FactoryAI is really interesting (aka “droid” on the CLI). if you’re into one‑shotting projects, you should definitely check it out. right now i have opus + gpt‑5.4 collaborating as orchestrator/worker/validator agents, all working together to refactor a typescript project. it’s been running for 6+ hours. really curious why it takes that long and burns 30M+ tokens. hoping the results will amaze me, because i already spent all my credits in the first hour. now i’m using my codex sub and keeping the droid subs as the orchestrator only. will update this tweet with the results!

Nemotron Super 3's strengths include: • Speed and efficiency: hybrid MoE architecture activating only 12B out of 120B total parameters, Nemotron 3 Super is fast and cost-effective, making it well-suited for high-throughput coding workflows • Reliable under real-world conditions: low error rates on core operations, and recovers well from build failures, missing dependencies, and unexpected environments rather than giving up • Consistent on procedural tasks: even standalone, Nemotron 3 Super reliably handles infrastructure setup, code modernization, and well-scoped algorithmic problems

Nemotron 3 Super is a cost-effective model that excels in coding when paired with a frontier planner. It follows detailed specifications reliably and executes multi-step coding tasks accurately. We are excited for the continued collaboration between the Factory and Nvidia research teams.

Over the weekend, I sent Droid on a Mission to create a home automation app. I wanted it to look like “xbox360 UI meets Spike Jonze's Her meets Liquid Glass” I told it find the devices on my network, loop me in when it needs help (IRL or w/ auth) and make a 3d mockup of my apt.

Alright, about to go take a nap. Let's see if @FactoryAI Missions are the new method. I made a comprehensive list of features I wanted built into a new app (from an old app I built), gave Droid the Kimi and GLM keys, and I'm about to enjoy a nap.