Learn to build the
solution you need.

A plain-words guide through the jungle of modern software and AI. Build something that solves a real problem you, and probably others, actually have. Understand every step, and share what you learn.

Come back and this column becomes your own reading list - what you were part-way through, and which of those pages changed since. Built in your browser, never sent anywhere.

What the checking found this month

Four things below. Two are corrections to pages on this site. Two are things most of the internet still has wrong. All four were found by opening a page, reading the vendor's current documentation next to it, and noticing they no longer agreed.

13 Nov 2026

Fable 5.1 was no longer Anthropic's top model, and its price was out of date

This page presented Claude Fable 5.1 at $10/$50 per MTok as Anthropic's most capable generally available model. Claude Fable 5.2 and Mythos 5.2 shipped on 9 November 2026 at $8/$40, with cache reads at $0.20, and the page was last updated on 28 October. Fable 5.1 is still Active and is now listed as the previous generation. The page also did not mention that fast mode on Opus 4.8 is removed on 30 November 2026.

Read the corrected page →
13 Nov 2026

Gemini 4 Carbon is available, and Gemini 4 Hydrogen now exists

This page said Gemini 4 Carbon was announced but not yet available to developers. It became generally available to paying developers on 10 November 2026, at the same introductory $2/$10 as Argon, rising to $4/$20 on 1 December. Gemini 4 Hydrogen, released on 12 November at a flat $1.50/$7.50 across its 1M context, was missing entirely, and Gemini 3.6 Flash's 19 November retirement was not in the price table.

Read the corrected page →
28 Oct 2026

Moonshot did release a model in October, two days before we said it had not

The October roundup said "nothing new found from Moonshot", based on a check of Moonshot's Hugging Face organisation on 25 October. Moonshot had announced Kimi K3.1 Pro and K3.1 Pro Highspeed on 23 October, describing K3.1 Pro as an open-weight model, with weights promised "by Nov 4" and API access from launch. Because the weights were not yet on Hugging Face, a check of Hugging Face alone could not find it. The entry now lists K3.1 Pro with its API prices and the unpublished licence. The lesson: a model hub shows weights, not releases. The lab's own blog and API changelog are the primary source for whether something shipped.

Read the corrected page →
28 Oct 2026

Thales did not measure a 53 percent bot share in 2023

This entry said Thales put bot traffic at 53 percent "as far back as 2023", citing a Fortune article. Thales's own releases say otherwise: its 2025 Bad Bot Report found automated traffic at 51 percent of all web traffic in 2024, "for the first time in a decade", and its 2026 report puts bots at over 53 percent. The entry also called Cloudflare's June 2026 figure the first time bots exceeded humans online; that is true only of Cloudflare's measure, requests for HTML pages, which differs from Thales's measure of all web traffic. We had relied on press coverage of the report instead of the report itself.

Read the corrected page →

564 of the 1868 pages here carry a last_verified date, and it is on the page itself, not in a changelog. If a claim matters to a decision you are making, check when it was last confirmed - and if it is stale, that is visible rather than hidden.

The current flagships, and what their spec sheets leave out

Context windows and prices are easy to find anywhere. The last column is the part that takes reading the release notes - and the part that changes what you should actually do.

ModelReleasedContext$/MTok in / outWhat the specs leave out
GPT-6 Astra3 Sep 20261,050,000$10 / $50The first model a lab has classified as crossing its own "Critical" cyber threshold. Prompts over 272k tokens bill at 2x input, 1.5x output.
Claude Fable 5.11 Sep 20261,000,000$10 / $50New tokenizer: roughly 30% more tokens for the same text, so 1M tokens is about 555,000 words, not 750,000. Since 22 September, Opus 5.5 claims roughly Fable-level results at $4 / $20.
Muse Spark 1.32 Sep 20261,000,000$1.25 / $4.25Meta's flagship, and not a Llama. Closed weights. A separate Contributor endpoint drops to $0.10 / $0.20 if you let Meta train on your prompts.
Nova 2 Lite2 Dec 20251,000,000$0.30 / $2.50The only generally available model in the Nova 2 family. Pro and Omni are still preview. That price is US East with global cross-region routing; geo routing costs 10% more, and Frankfurt about 30% more.
Mistral Large 32 Dec 2025256,000$0.50 / $1.50The cheapest frontier-class option here, and the only one whose weights you can download - Apache 2.0, including the flagship. Smallest context window of the five.

Prices are list rates and move often; every figure is dated on its source page. Full lineups for every vendor, including the open-weight and Chinese labs, in the 2026 LLM landscape.

Jump to any section

All 1868 pages, grouped: 239 guides, 256 tool pages, 114 comparisons, 497 glossary entries, and the rest. Open a group to see its sections - the same navigator you get on every page. For how the pieces fit together, see the five layers.

Open source

Tools built from the same experience.

Freelancer Templates and Freelancer Automation grew out of the same real-project work that powers this wiki. Both are free to use.