The LLM Network feed (/llms-network.txt)
The LLM Network feed at /llms-network.txt is the pipe between the Kariant catalog and external generative engines: a live, structured syndication corpus — with JSON, JSONL, and per-vertical shard variants — that engines consume to know what exists, what it does, and where to attribute it.
What the feed islink
llms-network.txt extends the llms.txt idea from “describe this site” to “syndicate this catalog”: a machine-readable feed of listings — names, categories, one-line summaries, canonical URLs, and each listing’s tier as a data field — published at the platform level and consumed by external engines. It is live today in four serializations: /llms-network.txt (markdown), /llms-network.json (JSON-LD ItemList), /feeds/catalog.jsonl (line-delimited JSON), and per-vertical shards at /llms-network/{software|ai-tools|services}.
How your listing gets inlink
Every published listing enters the feed automatically — all plans, no registration step. Tier is carried as a field on each record, not used as an inclusion gate. Plan tier changes the record’s richness: basic listings carry the compact record (name, summary, category, canonical URL, tier), while plan-stamped tiers (Kariant AI, Agentic AI) and platform-authored listings additionally carry pricing, real ratings when they exist, and links to the listing’s machine formats. What the feed says about you is drawn from your listing, so the same rule as everywhere applies: complete, current listings syndicate well; sparse ones syndicate sparsely.
Freshness and attributionlink
The feed reflects the catalog as it changes — update your listing and the syndicated record follows. Every entry carries its canonical listing URL, so when an engine cites you, attribution points back to your decision-ready detail page rather than a scraped fragment.
Frequently asked questions
Do I have to register my listing for the feed?expand_more
No registration step exists — every published listing syndicates automatically, on any plan. Keeping the listing complete and current is the only maintenance the feed needs from you.
What do paid tiers change if everyone is in the feed?expand_more
Payload richness. Basic records carry name, summary, category, canonical URL, and tier. Kariant AI and Agentic AI records add pricing, real ratings, and machine-format links — more substance for an engine to cite.
Which engines read the feed?expand_more
The feed is openly consumable by external generative and answer engines. Kariant publishes and maintains the pipe; it does not (and cannot) dictate each engine’s crawl schedule or usage.
Can I opt out?expand_more
Per-listing opt-out is not yet self-serve. Unpublishing a listing removes it from the feed immediately, and support can arrange exclusions in the meantime — support-assisted until the opt-out toggle ships in the Distribution console (kariant.app/partner/distribution), which already shows your listing’s real status on every surface.