About

About The Inference Directory

An independent, source-first directory to bring clarity to the fragmented inference market.

Our mission

Make the inference market comparable.

The inference market is fragmented. Two providers can look identical on a pricing page but differ completely on ownership, hosting region, model access, data residency, payment rails, KYC, and compliance. The directory exists to make those differences visible, sourced, and comparable.

Clarity

Cut through fragmented provider claims, inconsistent pricing pages, and unclear hosting language.

Trust

Source-backed records, explicit labels, and dated snapshots.

Actionable

Built for real deployment decisions: models, prices, access, hosting, and compliance in one record.

What we cover

Three directories. One standard.

Principles

What does not change.

  • Every claim links to a source.
  • The same criteria apply to every provider.
  • Labels are explicit: owned vs hosted, abliterated vs uncensored.
  • Benchmarks are never mixed into one ranking.
  • Every directory is a dated snapshot, not a live feed.
Who it's for

Built for people making a decision.

  • Developers choosing an inference API
  • Security researchers evaluating abliterated models
  • Procurement and platform teams comparing hosting and compliance
  • Founders mapping the inference market
Contact

Corrections, submissions, and questions.

If a provider has changed its catalog, pricing, ownership, or hosting, send the source. If you run a provider that belongs in one of the directories, send the catalog and endpoint documentation.

Independent. Source-first. Clearly sourced.

Browse directories