ModelPriceWatch.com
Last scan 2026-10-13 Models tracked 286 Providers 39 Cheapest paid Granite 4.0 H Micro $0.017/Mtok in Every price links to its source

Clef-flash

by Cloudflare · 9B parameters

Current new · 12d budget open weights cheap tier
Email me when Clef-flash gets cheaper
Availability: Hosted by Cloudflare on Workers AI as @cf/cloudflare/clef-flash, billed against the Workers AI neuron allowance (10,000 neurons a day free, then pay-as-you-go). The hosted model accepts up to 24,576 tokens; the open weights on Hugging Face (Apache 2.0) can be self-hosted.

Today's price · per 1M tokens

Input

$0.038

per 1M tokens

Output

$0

per 1M tokens

Blended

$0.028

blended $/1M — 3:1 weighted input:output

How this price is scoped: Cloudflare's own Workers AI pricing page lists @cf/cloudflare/clef-flash at $0.038 per M input tokens (3,455 neurons per M input tokens) and prints no output rate. Cloudflare's 2026-10-09 changelog announced the cut to this price and the smaller 24K-token hosted window that came with it; LiteLLM's public price table still shows the older $0.09 launch rate. Clef models do not charge for output tokens, because they return a decision with a probability for each allowed answer rather than generated text, so the $0 output here is Cloudflare's own published rate, not a missing number. Images and video are converted to input tokens and billed at the same input rate.

Source: official Cloudflare pricing · read MODELPRICEWATCH.COM · 2026-10-13

Price receipt

No confirmed receipt. We hold dated captures of Cloudflare's pricing page, but none of them tied Clef-flash to the price above on that page, so we do not claim this number is sourced. Check the provider's page directly before relying on it.

Overview

The fast, small member of Cloudflare's Clef family: a 9B multimodal decision model that reads a state (text, JSON, images or video) and typed questions and returns a probability for every allowed answer, not generated text. Cloudflare measured a 38.8 ms median latency at launch, about 13x faster than TypeSafe's Jev. On 2026-10-09 Cloudflare cut its price to $0.038 per 1M input tokens and shrank the hosted context window from 64K to 24K tokens to offer that price. No output charge.

Run it yourself

Deploy this open model on rented GPUs

Open weights mean you can self-host instead of paying per-token API prices. These platforms let you serve it on demand — often cheaper at scale.

RunPod →
Rent GPUs by the second and serve any open-weight model yourself.
Novita AI →
Serverless open-model inference billed per token — no GPU ops.
Vast.ai →
GPU marketplace — rent spot/on-demand GPUs cheaply and serve any open-weight model.

Partner links — we may earn a commission if you sign up, at no cost to you. We only list platforms we'd recommend regardless.

Capabilities

struck through = not supported
Input 3/4
Text ✓ Image ✓ Audio Video ✓
Output
Text ✓
Features 2/9
Prompt caching Reasoning Coding Fast inference Long context Open weights ✓ Multimodal ✓ Web search Realtime

Specifications

Provider
Cloudflare
Context window
25K tokens
Modality
text, image, video
Parameters
9B
Open source
Yes — open weights available
Released
Status
Current
Last updated
Tags
budgetmultimodalopen-weights

Availability verified: — listed on Cloudflare's own page

Official pricing page