Skip to content
Knight AI+AV Mark KNIGHT AI+AV
Home
Rankings
Start a project
Home / Rankings / AI Rankings
Knight Rating · Artificial Analysis source · Freshness shown

AI Rankings.
One rating. The whole field.

Source-attributed view of Artificial Analysis. The Knight Rating is their intelligence index normalized to 0–100; tied scores keep AA's native API order. Coding, speed, and cost are shown for context, not weighted. The response-reported observation or generation time and assessed freshness state appear below.

—
Models tracked
6
Categories
≤ 1h
Freshness window
—
Dataset generated
# Model Knight Rating Intel. Speed Cost Coding Δ
Loading Knight Rating board…
Source & Method

Sources, as reported.

The board renders the source records, scores, and status supplied by the Google rankings runtime. Open a source card to inspect its attributed URL; use the displayed observation or generation time and freshness state instead of assuming a refresh schedule.

Loading source…
How this runs

The Knight Rating method.

  1. 01

    Pull

    The page requests /api/rankings from the Google runtime for the selected model category. The response supplies source records and an observation or generation time; the browser assesses freshness without inventing an upstream cadence.

  2. 02

    Normalize

    The service returns a Knight Rating and any available intelligence, coding, speed, and cost axes. The page displays those values without silently inventing missing scores.

  3. 03

    Classify

    Every model is tagged open vs. closed and frontier vs. not, then deduped to a single canonical row.

  4. 04

    Rank

    Models sort by the field you select. The dataset's methodology metadata—not static page copy—is the authority for how each score was produced.

  5. 05

    Cache & serve

    If the response marks the snapshot stale—or its observation time exceeds the active freshness policy—the page labels it delayed. If the service is unavailable, the page shows a clear recovery state and retry instruction.

Use The Data

Open by default.

The full board is a public JSON endpoint under CC BY 4.0. Query it, cite it, or drop the live Top 10 onto your own site.

JSON API 
curl https://api.knightaiav.com/api/rankings?archetype=llm

Swap llm for text_to_image, text_to_video, image_editing, image_to_video, or text_to_speech. Per-model history lives at https://api.knightaiav.com/api/rankings-history?archetype=llm&id=<model-id>.

Embed widget 
<iframe src="https://knightaiav.com/embed/rankings.html"
        width="100%" height="560" loading="lazy"
        style="border:0;border-radius:12px;max-width:520px"></iframe>

A self-contained Top 10 view that requests the rankings endpoint and reports whatever dataset that endpoint returns. Preview the widget →