Weekly AI newsletter · Sent Sundays

Hype, itemized.

AI Receipts checks the week's AI claims against real numbers: prices, benchmarks and announcements. Sources linked, math shown, verdict given.

Subscribe free Read the issues

AI RECEIPTS

CLAIM CHECK · OCT 2026

"Argon is cheaper than GPT-6 Astra"

Argon, per index task$1.99

Astra, per index task$3.26

Argon output tokens/task~62,000

Astra output tokens/task~27,000

Argon at full price$3.98

VERDICT⚠️ PARTLY

True until Nov 30. Source: Artificial Analysis.

Latest issues

All issues →

#001 · Oct 18, 2026

Gemini 4 Argon is cheap. Until Nov 30

Gemini 4 Argon's 50% coupon itemized, OpenAI's 722 AI-written math papers audited, the White House's 'not optional' incident reporting, and what 2.2% of PNC households paying for AI really means.

Receipts

All claim checks →

⚠️ Partly

“OpenAI's AI solved hundreds of open math problems”

OpenAI released 722 AI-written math manuscripts. We itemized how many are machine-checked (162 in the Lean catalogue), peer-reviewed (0), and how many failures were disclosed (none).

Latest from the blog

All posts →

ExplainerOct 14, 2026

What is cost per task? Why the cheapest AI model per token isn't the cheapest per job

Per-token prices tell you the rate, not the bill. Cost per task counts every token a model burns to finish a job. Here is how it works, with real numbers from GPT-6, Claude and Gemini.