by CloudflareReleased 30 October 2026
This model takes in text and pictures, reads about 65 pages at one go, and hands back decisions as the typed fields you ask for. A buyer would use it to pull named figures out of a long file set, then work out totals and rough estimates from them. What sets it apart here is that it reads pictures alongside text in one pass, so a scanned form and its covering notes can be read together. It stops short of long drafting, since it gives back up to about 58 pages, so open-ended writing needs a different kind of model.
You pay for what each request uses. No subscription, no minimum.
One case, the ordinary use
Every rate
| Meter | ↑ you send | ↓ comes back |
|---|---|---|
| Text tokens | 33,28 DA |
per 1M tokens
One company runs this model. OpenDunes sends every request to it. DA per 1M tokens: ↑ what you send.
| Provider | ↑ Sent | Latency | Uptime |
|---|---|---|---|
| Cloudflare | 33,28 DA | … | … |
You pay the price of the provider that served your request. Prices read 15:18 · speed over the last 30 minutes · uptime over 3 days.
One HTTP call, with your questions named as you like.
import os, requests
BASE = "https://opendunes.com"
KEY = os.environ["OPENDUNES_API_KEY"]
H = {"Authorization": f"Bearer {KEY}"}
resp = requests.post(f"{BASE}/alpha/decisions", headers=H, json={
"model": "cloudflare/clairic",
"state": "The courier came two days late and the box was open.",
"questions": {
"is_complaint": {
"type": "noul",
"instructions": "Is it a complaint?",
"criteria": {
"true": "The customer is unhappy with the service.",
"false": "The customer is asking a question or saying thanks.",
},
},
},
})
print(resp.json())Runs this request once on your balance.
About 0,033 DA per case. ↑ 33,28 DA per 1M tokens.
No. You pay per use from your balance, with no subscription. Run it once from Call it in code, at the same price.
This model takes in text and pictures, reads about 65 pages at one go, and hands back decisions as the typed fields you ask for. A buyer would use it to pull named figures out of a long file set, then work out totals and rough estimates from them. What sets it apart here is that it reads pictures alongside text in one pass, so a scanned form and its covering notes can be read together. It stops short of long drafting, since it gives back up to about 58 pages, so open-ended writing needs a different kind of model.
Deposit dinars with your Edahabia or CIB card. Every model, Clairic included, is paid from that balance. No foreign card, no conversion.
Cloudflare names no languages for Clairic. Arabic, French and Darja: not stated.
Yes or no, one choice from your list, or a score on your scale. Each answer comes with its probability.
Cloudflare. Released 30 October 2026.
Clef Flash is the small, quick member of the Clef family, so it decides rather than writes. You send a situation as text, a table of data or up to four images, plus 1 to 64 questions that each have a set of allowed answers, and it returns the probability of each answer. Use it for quick, high-volume decisions such as routing a support ticket; it keeps about 24 pages in mind.
Clef does not write text. You send a situation as text, a table of data or an image, plus 1 to 64 questions of three kinds: yes or no, choose among options, or score against ranked criteria. It returns the probability of every allowed answer to every question, which suits routing a support ticket, and it keeps about 64 pages in mind.
Clef Omni is the member of the Clef family that can judge audio and video as well as text and images. You send a situation with up to four images, four audio clips of up to five minutes each, or two videos of up to a minute each, plus 1 to 64 questions that each have a set of allowed answers, and it returns the probability of each answer. It keeps about 64 pages in mind.
From what its maker and two benchmark sites publish, and where it stands on OpenDunes. Read 7 Nov 2026.
Pulling named figures from a bundle of contract pages and their scans
Totalling amounts across a stack of scanned invoices and handwritten notes