Skip to content
worth noting New models

Cloudflare has made Clairic available for extracting data from documents

only one source so far

Clairic is available in preview on the Workers AI platform. It processes the requested fields from a document in a single pass and returns values with confidence levels; for missing data, it returns null.

Cloudflare has made Clairic available in preview on the Workers AI platform. The user specifies the requested fields, such as an order number or the total invoice amount. The model processes them in a single pass and returns values in the requested format along with confidence levels. If data is missing, it returns null.

Cloudflare demonstrates its use in claims handling: Clairic extracts data from a message and estimates the repair cost, Clef chooses between a repair and a refund based on company rules, and GLM-5.3 writes a response to the customer. According to Cloudflare, this setup completes the task in 19.3 seconds, while GLM-5.3 alone takes 61.7 seconds and costs almost twice as much. This is a comparison reported by the company for this specific workflow.

According to Cloudflare, Clairic can also handle simple calculations, estimates and forecasts, but does not yet reliably perform longer chains of calculations. The company recommends keeping those in the application's own code. The release of open weights is expected to follow soon.

What changed

Why it matters

When processing documents, developers can separate the extraction of specific data from decisions based on rules and the writing of responses. The example shows potential time and cost savings in claims handling, but the results come from Cloudflare and do not establish the same benefits for other tasks.

Release card

Clairic

Cloudflare

in the preview
Inputs
The model processes documents and requested fields. It returns the corresponding values in the requested format along with confidence levels; when data is missing, it returns null.
Availability
The model is available in preview on the Workers AI platform. According to the article, the weights are scheduled for release later.
According to the sources, it is suitable for
  • Extracting order numbers and total amounts from invoices.
  • Extracting data from claims and estimating repair costs.
  • Simple calculations, estimates and forecasts.
Documented limits
  • According to Cloudflare, it does not yet reliably perform longer chains of calculations.

Cloudflare presents Clairic as a faster and cheaper tool for extracting data from documents than large language models. It processes the requested fields in a single pass.

The card summarizes information from the article and any dated corrections, with a link to the original source. It is not our assessment of the model. It does not yet have a dedicated editorial profile. Model selection and other announcements →

Two audiences, two different impacts

What this means

01

For individuals

A developer can specify a list of fields to extract and work with the values, confidence levels and null indicators for missing data. According to the company recommendation, longer sequences of calculations must remain in the developer's own code.

What to do Try Clairic in preview on a sample document that is missing one of the requested fields, and check the returned values and confidence levels.
More practical updates →
02

For a business

Companies automating claims handling can test splitting the process into data extraction, decisions based on company rules and response generation. Cloudflare reports shorter processing times and lower costs for this approach than for using GLM-5.3 alone.

Processes
What to decide Use an anonymized sample of claims to compare the accuracy, processing time and costs of the proposed setup with the current process.
More business impacts →
Clairic Cloudflare Workers AI

Check the original

Event sources

only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.

1
The Decoder (daily AI news) independent context · first detected Cloudflare's Clairic pulls data from documents faster and cheaper than LLMs