Cloudflare (Workers AI) · Decision model · Last verified 2026-10-15

Cloudflare Clef (Clef, Clef-flash, Clef-omni)

Verified

Cloudflare-hosted decision models on Workers AI that take a state and a schema of typed questions (noul, choice, score) and return a probability for every allowed option. Three variants: Clef (27B), Clef-flash (9B), Clef-omni (30B MoE, 3B active).

What it does

  • Model IDs: @cf/cloudflare/clef, @cf/cloudflare/clef-flash, @cf/cloudflare/clef-omni; System One-style request body via Workers AI binding or REST.
  • State can be text, JSON, images, or video; Clef-omni also accepts audio. Images are embedded (no remote URLs).

Pricing

  • Clef $0.240, Clef-omni $0.150, Clef-flash $0.038 per 1M input tokens; no output-token charge. Media are converted to input tokens. [Verified 2026-10-15 on Cloudflare Workers AI pricing page (last updated Oct 9, 2026) and model pages]
  • Workers AI billing: $0.011 per 1,000 Neurons; 10,000 Neurons/day free on Free and Paid plans. [Verified 2026-10-15 against official page]

Limits

  • Context: Clef 65,536 tokens; Clef-flash 24,576; Clef-omni 64,000. [Verified 2026-10-15 on Cloudflare model pages]
  • Clef-omni: 1–64 questions; up to 4 images (4 MiB each), 4 audio clips, 2 videos (60 s each). Each image capped at 1,024 tokens. [Verified 2026-10-15 on clef-omni model page]

Status

  • Listed on Workers AI model catalog and pricing page. [Verified 2026-10-15 against official page]

Safety & fit notes

  • Cheapest per-token hosted option here (Clef-flash) but with the smallest context; long text state is truncated to fit.
  • Test calibration and order sensitivity on your own labels; vendor pages publish no accuracy figure.

Compare

Sources