# What machines read when they visit an AI tool directory (slim) > A week of Anchor Terminal traffic in charts. Comparisons, execution sandboxes, decision APIs and voice cloning, with the limits of what a server log can tell us. - Full: https://www.anchorterminal.com/blog/what-machines-read.md (~2,850 tokens) · this version ~1,680 tokens · JSON https://www.anchorterminal.com/blog/what-machines-read.json · canonical https://www.anchorterminal.com/blog/what-machines-read - Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · Updated: 2026-10-13 By Anchor Terminal research, 13 October 2026. 1,117 words, about 5 min to read. The most interesting part of our weekly report is where the traffic went. We label browser traffic by the browser a request claims to come from. - 577,898: successful responses and redirects - 515: listings added to the directory The reporting window, 6–12 October 2026. Source [1]. The more useful story is what was requested. ## Comparisons drew the automated traffic Head-to-head pages drew more requests than individual listings from both agents and crawlers. Comparing tools drew more requests than reading individual listings | Page type | Agents | Crawlers | | --- | --- | --- | | Head-to-heads | 11412 | 175717 | | Individual listings | 6008 | 93420 | Each series has a zero-based scale. A request cannot tell us whether a client was choosing a product, collecting material for a future answer or building an index. The leading comparison in agent-classified traffic was [Serper versus You.com](/compare/serper-vs-you-com-api) [[1]](#ref-1). [Serper](/tools/serper) exposes Google search results through an API, including news, images and maps [[2]](#ref-2). The choice involves how much of the search-and-read workflow to assemble, how an agent gets access and where a person must intervene. ## The busiest days tell different stories Browser traffic rose towards the end of the reporting window. Daily requests answered successfully or redirected | Day | Browser | Agents | Crawlers | | --- | --- | --- | --- | | Tue 6 | 8094 | 5481 | 35884 | | Wed 7 | 33176 | 1805 | 30201 | | Thu 8 | 9732 | 4431 | 70892 | | Fri 9 | 5246 | 6874 | 82370 | | Sat 10 | 14582 | 4175 | 63269 | | Sun 11 | 40349 | 2910 | 56189 | | Mon 12 | 43540 | 3860 | 53739 | Each series has a zero-based scale. We are treating this as a baseline. ## New listings brought execution and identity into view The catalogue grew during the week, but an addition to Anchor Terminal is not necessarily a newly launched product. Most-viewed additions among browser traffic | New listing | Views since listing, within the week | | --- | --- | | Microsoft Execution Containers | 31 | | Gmail API | 29 | | AgentCore Memory | 28 | | AgentCore Identity | 26 | | Amazon Ads API | 25 | | Amazon Nova Embeddings | 25 | Each series has a zero-based scale. Microsoft Execution Containers runs untrusted code, including model-generated code, inside a configurable sandbox. Its first stable SDK release arrived during the reporting window, giving developers a versioned surface across Rust, .NET and Node.js [[5]](#ref-5). [AgentCore Identity](/tools/agentcore-identity) addresses another part of the same problem. These are the operational questions that appear when an agent can take actions. ## Agents also read about making decisions [OpenAI Decisions](/tools/openai-decisions-api) led the new-listing table for agents. Most-viewed additions among agents | New listing | Views since listing, within the week | | --- | --- | | OpenAI Decisions API | 24 | | Google Sheets API | 12 | | WhatsApp Cloud API | 12 | | Seedance on BytePlus ModelArk | 10 | | Meta Marketing API | 9 | Each series has a zero-based scale. The Decisions endpoint evaluates text and images and returns a probability, a choice from supplied options or a score against ordered levels. An application choosing which department should receive a ticket needs a bounded answer it can act on. ## Voice cloning has a workflow question [Cartesia versus ElevenLabs voice cloning](/compare/cartesia-voice-cloning-vs-elevenlabs-voice-cloning) led the browser comparison table [[1]](#ref-1). The leading head-to-heads among browser traffic | Comparison | Views | | --- | --- | | Cartesia vs ElevenLabs voice cloning | 84 | | Cloudflare Clef vs Convai Laya | 70 | | Cloudflare Clef vs TypeSafe Jev | 44 | | Infobip Calls vs Voximplant | 27 | | Attio vs Salesforce | 24 | Each series has a zero-based scale. Cartesia's instant cloning route starts from a short recording, while its professional route trains on substantially more audio [[8]](#ref-8). These are different cloning modes, rather than a direct quality comparison. ## Markdown and JSON are being read Our head-to-head pages were retrieved in several formats. Formats requested by agents on head-to-head pages | Format | Share of requests, rounded per cent | | --- | --- | | Markdown | 43 | | HTML | 37 | | JSON | 21 | Each series has a zero-based scale. The rounded shares do not add up exactly. ## The blog had a quieter week Decision models also led the blog's browser traffic, followed by why we built the directory and our reading of vendors' data terms. Blog posts read by browsers | Post | Views across formats | | --- | --- | | Everyone wants your decisions | 33 | | Why we built Anchor Terminal | 27 | | What the terms say about your data | 17 | | Personal Agent Protocol, first draft | 12 | | Train your own decision model | 8 | | Agent tools and enterprise readiness | 7 | | Week 41 review | 3 | Each series has a zero-based scale. These are subjects we can keep investigating in the listings and explain more fully on the blog. ## How to read these charts The figures are frozen to the reporting window, rather than live counters. Browser, agents and crawlers are inferred from User-Agent strings. --- - Older: [Decision models, part two: train your own](https://www.anchorterminal.com/blog/decision-models-train-your-own.md) - Newer: [Cypherpunks write code, and now agents read it](https://www.anchorterminal.com/blog/rea-cypherpunks-write-code.md) - All posts: https://www.anchorterminal.com/blog/index.md ## References Product sources read on 13 October 2026. Traffic comes from the supplied weekly report; raw logs are not published. [1] Anchor Terminal weekly session report, 6–12 October 2026, not published. [2] Serper, product page. [3] You.com, MCP server documentation. [4] Microsoft, MXC repository. [5] Microsoft, MXC SDK v1.0.0 release notes, October 2026. Source for the stable SDK release and supported languages. [6] AWS, AgentCore Identity documentation. [7] OpenAI, Decisions guide. [8] Cartesia, Instant Voice Clone guide. [9] ElevenLabs, Professional Voice Cloning guide.