Faharas — The AI Index · About

Making the AI ecosystem
legible.

Faharas is the AI Index — a structured, verified reference for everything artificial intelligence. This page is about why we built it, how we keep it honest, and what comes next.

The story

The AI ecosystem is the fastest-moving field in computing. Every week brings a new model, a new tool, a new benchmark — and with them, a flood of pages, posts, and directories that often disagree. Choosing a model or a tool means stitching facts together from a dozen sources, most of them written to sell you something, many of them already out of date.

Faharas starts from a simpler question: what if every important resource had one canonical record — one place where its identity, its facts, its relationships, and its history all live? Not a directory of links. Not a feed of announcements. A reference work for the field, maintained like one.

That is the whole idea behind the index: one record per resource, facts tied to their sources, and change kept with history — so that searching, comparing, testing, and working with AI starts from a single, trustworthy place.

Faharas itself has history too — a decade that began as an Arabic content community in 2019, survived two shutdowns, and is now rebuilt around this idea. Read the full story

How we keep it honest

An index is only as good as its discipline. These are the editorial standards every record is held to.

Evidence over claims

Every sourced fact — a release date, a price, a license — carries its source. What we calculate ourselves is always labeled as our own measure, never presented as a provider's claim.

Nothing invented

If a field is unverified or unknown, we say so. The interface shows "unavailable" or "unverified" instead of a guess — missing data is never fabricated to fill a page.

Nothing disappears

Retired, merged, and renamed resources keep their history — with redirects, aliases, and provenance — so an old article or a saved record stays understandable.

Community, moderated

Corrections and contributions are welcome from anyone. They pass an explicit review before they can touch canonical data — and private data never leaks into public records.

How the index works

Every resource in Faharas moves through the same disciplined pipeline — once.

Identify

One canonical identity per resource — names, aliases, former names, and external identifiers, so identity survives renames and mergers.

Classify

Every resource gets one primary type, a category, and tags — consistent enough to browse, flexible enough to describe reality.

Verify

Facts are tied to sources and given verification states, so everything on a page can be traced back to where it came from.

Relate

Explicit relationships connect the ecosystem: who develops what, what uses what, what evaluates what — never left to guesswork in free text.

Publish

One record, many surfaces. Search, directories, comparisons, and your library all read the same canonical record — never a copy.

What's inside

The index spans the full ecosystem — from the models themselves to the knowledge around them:

AI Models Companies & Labs AI Tools AI Agents MCP Servers APIs & SDKs Frameworks & Libraries Prompts & Skills Datasets Papers & Courses Benchmarks & Comparisons Leaderboards & Updates

Every one of them — one record, verified, related, with history.

What comes next

We are building in phases — the platform grows in public, one capability at a time.

The index opens

On 14 November 2026: models, companies, tools, agents, MCP servers, APIs, and frameworks — with search, comparisons, and verification states from day one.

Knowledge & libraries

Prompts, skills, datasets, papers, guides, and benchmarks — plus the Playground, to test supported models directly.

The personal platform

Libraries, collections, workspaces, and AI connections — the tools that turn the index into a place you work in.

We open on
14 November 2026.

Faharas — The AI Index

The first phase goes live then — and it is only the beginning.

You're on the list. One email when we open — nothing else.

Back to the announcement

Making the AI ecosystemlegible.