computed from published catalog judgments · evidence levels shown below

The 2026 AI Replaceability Report

DeepFeather publishes an editorial judgment for each product: can AI replace its core workflow? This report aggregates those judgments. A verdict is not proof of a completed build; the evidence section below shows exactly how many entries have been built, curated, or only estimated.

products tracked
1,025
catalog entries, each with a verdict
verdict: yes
16%
159 of 1,025 products
China-market apps
48
listed for the Chinese market

Evidence behind the verdicts

Verdicts and build evidence are separate. Only Built & verified means DeepFeather has a dated, reviewable build result.

Built & verified0dated, reviewable build evidence · 0%
Build failed0a documented build attempt did not pass · 0%
Curated prompt128editor-reviewed plan or prompt; not a completed build · 12%
Catalog estimate897judgment without verified build evidence · 88%

Verdict distribution

YES159editorial judgment: AI can likely replace the core workflow · 16%
KINDA455core loop yes, but a real moat remains · 44%
NOT REALLY411AI is not close yet · 40%

Bills represented by the verdicts

712 of 1,025 products publish a monthly price. Restricting to the 456 with a yes or kinda verdict, listed sticker prices add up to $9,734/mo — $116,808/yr. These are listed bill totals, not savings estimates and not DeepFeather revenue.

Top categories

Coverage

48 apps serve the Chinese market, each with local pricing and notes.

298 entries have complete Chinese content and live under /zh/.

Start with these

Explore the data:the full list ·all categories ·live site analytics