agi.tech is built with Framer, hosted on DigitalOcean behind Cloudflare. It sends email through Amazon SES.
| Technology | Version | First seen |
|---|---|---|
| CMS & Site Builders1 | ||
| – | 9 Nov 2026 | |
| CDN & Hosting2 | ||
| – | 9 Nov 2026 | |
| – | 9 Nov 2026 | |
| Transactional Email2 | ||
| – | 9 Nov 2026 | |
| Fonts & Icon Fonts1 | ||
| – | 9 Nov 2026 | |
| Translation & Localization1 | ||
| – | 9 Nov 2026 | |
Registration from the registry's RDAP record, the certificate the site serves, and its rank in public traffic lists.
The crew hasn’t read this site’s colors yet. We read a palette when we capture a site’s screenshot, busiest sites first. Meanwhile, browse palettes from other sites.
Counted from every stack change we've recorded: of the sites we watched running one of this site's tools, how many went on to add another. A pattern across sites, not a forecast for this one.
From what the site says about itself, and account IDs it shares with other sites.
smtp.google.commargo.ns.cloudflare.comalexandra.ns.cloudflare.comgoogle-site-verification=hIVTeWv8mXRzieZqkh3ImNu-MWKQc57d3JntSRu6DqQgoogle-site-verification=VdOiiZSlJBRJ_sP7XTzMl_IStyLRkpcLQW15S20aLKMv=spf1 include:_spf.google.com ~alltwilio-domain-verification=83c7b9135057ce538a4bc6c5a9ef4631v=DMARC1; p=none; rua=mailto:9b9a48a96b91467e8fe55d5b60c92170@dmarc-reports.cloudflare.netUser-agent: OAI-SearchBot User-agent: ChatGPT-User User-agent: Claude-SearchBot User-agent: Claude-User User-agent: PerplexityBot User-agent: Perplexity-User User-agent: Googlebot User-agent: Bingbot User-agent: Applebot User-agent: DuckAssistBot Allow: / # Training crawlers: allowed on purpose. A site that wants to be known by # future models should stay in the training corpus. Revisit if legal objects. User-agent: GPTBot User-agent: ClaudeBot User-agent: Google-Extended User-agent: CCBot Allow: / User-agent: * Allow: / Sitemap: https://theagi.company/sitemap.xml
# AGI, Inc. > AGI, Inc. is a San Francisco research lab building action AI: AGI-0, a family of small action models (2B to 20B+ parameters); AGI Agent, an assistant that sees the screen and acts across any app on Android with no APIs or integrations; and AGI-RUN, an on-device inference engine for phones, laptops and edge hardware. ## About - [About AGI, Inc.](https://theagi.company/about): what the company builds, the team, investors (Menlo Ventures, Point72) and frequently asked questions. - [Contact](https://theagi.company/contact): partnerships, OEM and device integrations, press and careers. ## Products - [AGI-0](https://theagi.company/agi-0): small action models that see the screen, reason and act across any app; deployable in the cloud, on-prem or at the edge. - [AGI app on Google Play](https://play.google.com/store/apps/details?id=com.agiinc.androidagent): the Android agent that completes tasks across apps on your phone. ## Announcements - [Collaboration with Qualcomm Technologies](https://theagi.company/blog/agi-inc-announces-collaboration-with-qualcomm-to-bring-agentic-ai-to-snapdragon-powered-devices): agentic AI for Snapdragon-powered devices, announced March 2026. - [MWC 2026 demo with a proof-of-concept Lenovo integration](https://theagi.company/blog/agi-inc-demonstrates-prototype-mobile-use-agent-platform-at-mwc-with-proof-of-concept-lenovo-integration): the mobile-use agent platform shown at Mobile World Congress, March 2026. - [Rich Miner, co-founder of Android, joins as advisor](https://theagi.company/blog/rich-miner-joins-agi-as-advisor): August 2026. - [Partnership with Visa for agentic commerce](https://theagi.company/blog/agi-inc-visa): August 2025. ## Research - [AndroidWorld result](https://theagi.company/blog/android-world): 97.4% task success on the AndroidWorld mobile-agent benchmark, October 2025. - [OSWorld result](https://theagi.company/blog/osworld): 76.26% task success on the OSWorld computer-use benchmark, October 2025. - [REAL Bench](https://theagi.company/blog/introducing-real-bench): a web-agent evaluation benchmark built on replicas of real sites. - [AGI-RUN benchmarks](https://theagi.company/blog/agi-run-benchmarking-agis-on-device-inference-engine): the on-device inference engine measured against the field, up to 9.9x faster than state-of-the-art engines. - [Processing-near-memory](https://theagi.company/blog/processing-near-memory): up to 96% faster on-device inference without hardware changes. ## Events - [Edge Cases](https://theagi.company/edge-cases): AGI, Inc.'s research night for on-device AI in San Francisco. One researcher presents to 15 to 25 people, followed by off-the-record Q&A and dinner. Nothing is recorded. Past speakers came from Glass Imaging, NVIDIA and AGI, Inc. Researchers working on on-device inference, small models, efficient training, accelerators and agents on devices can apply to present. ## Glossary - [Glossary](https://theagi.company/glossary): definitions of the terms behind AI agents that operate phones and apps - [Mobile use](https://theagi.company/glossary/mobile-use): Mobile use is an AI agent's ability to operate the apps on a phone or tablet and finish tasks for a person. - [Phone use](https://theagi.company/glossary/phone-use): Phone use is an AI agent using the apps on a phone to finish tasks, often on the person's own phone and accounts. - [Agentic OS](https://theagi.company/glossary/agentic-os): An agentic OS is an operating system with an AI agent that finishes tasks across the device's apps when the user asks. - [Computer use](https://theagi.company/glossary/computer-use): Computer use is an AI agent operating a computer's apps and websites through the screen, the way a person does. - [Agentic commerce](https://theagi.company/glossary/agentic-commerce): Agentic commerce is shopping that an AI agent completes for a person, from finding the item to checkout, with the person approving the payment. - [Agent layer](https://theagi.company/glossary/agent-layer): An agent layer is the part of a device's software that turns a person's request into finished tasks across the device's apps. - [Agent harness](https://theagi.company/glossary/agent-harness): An agent harness is the software around an AI model that runs a task from start to finish and keeps the person in control. - [Accessibility service (Android)](https://theagi.company/glossary/accessibility-service): An accessibility service is an Android feature that lets an app you approve read the screen and act in other apps on your behalf. - [MCP (Model Context Protocol)](https://theagi.company/glossary/mcp): MCP (Model Context Protocol) is an open standard that lets AI assistants and agents connect to outside tools and data. - [Action model](https://theagi.company/glossary/action-model): An action model is an AI model that returns the next action in an app or on a screen instead of a written answer. - [On-device action model](https://theagi.company/glossary/on-device-action-model): An on-device action model is an action model small enough to run on a phone's or PC's own chip. - [Small action model (SAM)](https://theagi.company/glossary/small-action-model): A small action model (SAM) is an action model compact enough to run on a phone, PC or other edge device. - [Vision-language-action (VLA) model](https://theagi.company/glossary/vision-language-action-model): A vision-language-action (VLA) model takes an image and an instruction and returns an action, for a robot or on a screen. - [On-device inference](https://theagi.company/glossary/on-device-inference): On-device inference means running a trained AI model on the device itself, such as a phone or PC, instead of on a server. - [Hybrid inference](https://theagi.company/glossary/hybrid-inference): Hybrid inference runs an AI model on the device where the hardware allows and sends heavier steps to the cloud or on-prem servers. - [Inference engine](https://theagi.company/glossary/inference-engine): An inference engine is the software that runs a trained AI model on the chips a device has. - [NPU (neural processing unit)](https://theagi.company/glossary/npu): An NPU (neural processing unit) is the part of a chip built to run AI models quickly and at low power. - [Prefill throughput](https://theagi.company/glossary/prefill-throughput): Prefill throughput is how fast an AI model reads its input before it starts writing, measured in tokens per second. - [AndroidWorld](https://theagi.company/glossary/androidworld): AndroidWorld is a public benchmark that tests AI agents on 116 tasks across 20 real Android apps. - [OSWorld](https://theagi.company/glossary/osworld): OSWorld is a public benchmark that tests computer-use agents on 369 tasks on real computers, across desktop apps, files and websites. ## Developer - [Documentation](https://docs.agi.tech): developer docs for building on AGI. - [An MCP for every app on your phone](https://theagi.company/blog/agi-mcp-for-every-app-on-your-phone): how agents reach apps that have no API. - [Blog](https://theagi.company/blog): research, product updates and press coverage.
| – |
| 9 Nov 2026 |