Cloudflare · Oct 19, 2026
Cloudflare previews Clairic, a fast model for extracting and estimating typed values
Markdown · JSON · Original source
Cloudflare announces Clairic, a preview model that returns structured values—such as names, dates, totals, estimates and forecasts—from text or images. Unlike a chat model that generates free-form text, Clairic returns answers in specified types, with confidence information and, for estimates and forecasts, uncertainty intervals. Cloudflare says the model is designed for single-pass processing and reports lower latency and cost than using an LLM for certain extraction tasks. In its own comparison on a damaged-product claim workflow, Cloudflare reports that a Clairic–Clef–GLM pipeline took 19.3 seconds and cost 49% less than an all-GLM pipeline; this is a company-reported test, not evidence of performance across other workflows. The preview is hosted on Workers AI, and Cloudflare says it plans to release open weights. The company cautions that the early model can struggle with longer arithmetic, aggregation and unfamiliar data; confidence and well-formed outputs do not guarantee correctness.
Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.
Original source
Cloudflare Blog · Article
Source details
Original source date: Oct 19, 2026 (2026-10-19T17:48:26.614Z).
Date basis: Date stated by the publisher.
Added to Escape Meridian: 2026-10-19T19:19:07.010Z. First observed: 2026-10-19T17:51:10.245Z. Retrieved: 2026-10-19T18:01:09.391Z.
Referenced by
Recorded citations and public file attachments. A reference alone does not establish agreement or influence.
No public references recorded here yet.
Limits
Performance and cost comparisons are reported by Cloudflare for its own example workflow; the evidence does not establish results across other tasks or deployments. The source describes Clairic as an early preview checkpoint and says forecasting is basic and not a replacement for specialized time-series models. Confidence scores and uncertainty intervals do not guarantee correctness, and Cloudflare says intervals may be too narrow on unfamiliar data.
Open question
How well will Clairic’s confidence scores and uncertainty intervals hold up on unfamiliar data and workflows beyond Cloudflare’s reported examples?
Source-stated
Cloudflare announced Clairic as a hosted preview model that extracts, computes, estimates and forecasts values from unstructured inputs; it says open weights are planned.
Our first one, Clairic, is available hosted on Workers AI as @cf/cloudflare/clairic.We intend to release the model open-weights soon.
Source-stated
Clairic returns values in requested data types, rather than generating free-form text, and provides confidence information; estimates and forecasts can include uncertainty intervals.
Every scalar answer comes with p_null (the probability that the value isn’t in the inputs) and confidence. Every estimate answer comes with a median and an 80% interval
Source-stated
Cloudflare reports that in its damaged-product claim comparison, a pipeline using Clairic, Clef and GLM-5.3 took 19.3 seconds versus 61.7 seconds for an all-GLM pipeline, and cost 49% less.
finished in 19.3 seconds against 61.7 seconds: 69% less time, or about 3.2 times faster. Clairic and Clef handled extraction and the decision in under a second combined, where GLM-5.3 took 38 seconds for the same two steps. It also costs 49% less to use the cascaded pipeline vs the all-LLM pipeline.
Source-stated
Cloudflare says the preview model can struggle with longer chains of arithmetic and aggregation across many table rows, and warns that calibrated intervals are not guaranteed on unfamiliar data.
Calibration is measured, not guaranteed. Intervals are trained to be calibrated and are on most held-out sets, but they can run narrow on unfamiliar data.
Source-stated
Cloudflare emphasizes that typed outputs are not necessarily correct, so users should use confidence information and additional review when deciding what to automate.
The architecture makes every answer well formed, not necessarily correct. Use confidence, p_null and the intervals to decide what to automate (in an appropriate workflow context) and what to send off for additional review.
Source and contributor context
Cloudflare · Publisher
The publisher is identified here. No individual author has been inferred.
Source media
Media declared by the original publisher; availability checked separately from the editorial reading.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
Large language model
A language model with many learned numerical settings, used to process and generate language.
Time-series
Observations arranged over time, such as monthly sales; forecasting uses their patterns to estimate later values.
Cloudflare
Cloudflare provides networking, security and computing services that help run websites and applications.
Editorial provenance
Version 1. Publication ID: 98f6223d-7058-4097-83f9-72a986caa2e6.
Prepared by server-source-span-replacement/v1 (original draft: gpt-6-luna); grounding review by gpt-6-luna.
Source check: passed by anthropic/claude-opus-5.5 on 2026-10-19T19:19:07.010Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: material_development.
Preparation correction preparation-recovery-clairic-20261019-v1: exact source spans replaced an invalid excerpt using preparation-source-spans-v1. Original failed attempt ba3efa223f6f6917c102b65f684ec9ee8fbb8487ba1143610c8fd9353b57fb43 and its reservation are retained. Narrative wording is unchanged; the full corrected draft passed separate grounding review and source audit.
Observation: 5319a9e6c8d552681b48bc522cb25d190087c59659a1b4bb4b653cf606b117b5. Source text hash: d68807e08d2cfd5f9bb4081b7040dd54ba933fcb12dd15db696cda01a9c613a5.
Revision and source-check history
Version 1 · 2026-10-19T19:19:07.010Z
Cloudflare previews Clairic, a fast model for extracting and estimating typed values
Cloudflare announces Clairic, a preview model that returns structured values—such as names, dates, totals, estimates and forecasts—from text or images. Unlike a chat model that generates free-form text, Clairic returns answers in specified types, with confidence information and, for estimates and forecasts, uncertainty intervals. Cloudflare says the model is designed for single-pass processing and reports lower latency and cost than using an LLM for certain extraction tasks. In its own comparison on a damaged-product claim workflow, Cloudflare reports that a Clairic–Clef–GLM pipeline took 19.3 seconds and cost 49% less than an all-GLM pipeline; this is a company-reported test, not evidence of performance across other workflows. The preview is hosted on Workers AI, and Cloudflare says it plans to release open weights. The company cautions that the early model can struggle with longer arithmetic, aggregation and unfamiliar data; confidence and well-formed outputs do not guarantee correctness.
Original source
Cloudflare Blog · Article
Source details
Original source date: Oct 19, 2026 (2026-10-19T17:48:26.614Z).
Date basis: Date stated by the publisher.
Added to Escape Meridian: 2026-10-19T19:19:07.010Z. First observed: 2026-10-19T17:51:10.245Z. Retrieved: 2026-10-19T18:01:09.391Z.
Limits
Performance and cost comparisons are reported by Cloudflare for its own example workflow; the evidence does not establish results across other tasks or deployments. The source describes Clairic as an early preview checkpoint and says forecasting is basic and not a replacement for specialized time-series models. Confidence scores and uncertainty intervals do not guarantee correctness, and Cloudflare says intervals may be too narrow on unfamiliar data.
Open question
How well will Clairic’s confidence scores and uncertainty intervals hold up on unfamiliar data and workflows beyond Cloudflare’s reported examples?
Source-stated
Cloudflare announced Clairic as a hosted preview model that extracts, computes, estimates and forecasts values from unstructured inputs; it says open weights are planned.
Our first one, Clairic, is available hosted on Workers AI as @cf/cloudflare/clairic.We intend to release the model open-weights soon.
Source-stated
Clairic returns values in requested data types, rather than generating free-form text, and provides confidence information; estimates and forecasts can include uncertainty intervals.
Every scalar answer comes with p_null (the probability that the value isn’t in the inputs) and confidence. Every estimate answer comes with a median and an 80% interval
Source-stated
Cloudflare reports that in its damaged-product claim comparison, a pipeline using Clairic, Clef and GLM-5.3 took 19.3 seconds versus 61.7 seconds for an all-GLM pipeline, and cost 49% less.
finished in 19.3 seconds against 61.7 seconds: 69% less time, or about 3.2 times faster. Clairic and Clef handled extraction and the decision in under a second combined, where GLM-5.3 took 38 seconds for the same two steps. It also costs 49% less to use the cascaded pipeline vs the all-LLM pipeline.
Source-stated
Cloudflare says the preview model can struggle with longer chains of arithmetic and aggregation across many table rows, and warns that calibrated intervals are not guaranteed on unfamiliar data.
Calibration is measured, not guaranteed. Intervals are trained to be calibrated and are on most held-out sets, but they can run narrow on unfamiliar data.
Source-stated
Cloudflare emphasizes that typed outputs are not necessarily correct, so users should use confidence information and additional review when deciding what to automate.
The architecture makes every answer well formed, not necessarily correct. Use confidence, p_null and the intervals to decide what to automate (in an appropriate workflow context) and what to send off for additional review.
Source and contributor context
Cloudflare · Publisher
The publisher is identified here. No individual author has been inferred.
Source media
Media declared by the original publisher; availability checked separately from the editorial reading.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
Large language model
A language model with many learned numerical settings, used to process and generate language.
Time-series
Observations arranged over time, such as monthly sales; forecasting uses their patterns to estimate later values.
Cloudflare
Cloudflare provides networking, security and computing services that help run websites and applications.
Editorial provenance
Version 1. Publication ID: 98f6223d-7058-4097-83f9-72a986caa2e6.
Prepared by server-source-span-replacement/v1 (original draft: gpt-6-luna); grounding review by gpt-6-luna.
Source check: passed by anthropic/claude-opus-5.5 on 2026-10-19T19:19:07.010Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: material_development.
Preparation correction preparation-recovery-clairic-20261019-v1: exact source spans replaced an invalid excerpt using preparation-source-spans-v1. Original failed attempt ba3efa223f6f6917c102b65f684ec9ee8fbb8487ba1143610c8fd9353b57fb43 and its reservation are retained. Narrative wording is unchanged; the full corrected draft passed separate grounding review and source audit.
Observation: 5319a9e6c8d552681b48bc522cb25d190087c59659a1b4bb4b653cf606b117b5. Source text hash: d68807e08d2cfd5f9bb4081b7040dd54ba933fcb12dd15db696cda01a9c613a5.
passed · 2026-10-19T19:19:07.010Z · anthropic/claude-opus-5.5
Publication: 98f6223d-7058-4097-83f9-72a986caa2e6
attribution — pass: The draft correctly attributes the announcement, model description, comparison figures and caveats to Cloudflare, the publisher of this blog post. Company-reported results are framed as Cloudflare's claims.
Today, we're announcing a preview of a new class of System One-style models: extraction models. Our first one, Clairic, is available hosted on Workers AI
summary — pass: The title, summary and question faithfully convey the development: a preview model that returns typed values from text or images, with confidence and intervals. They also cover the company-run comparison and its caveats. Clef and GLM are not explained in the summary, but the main development stays understandable. The question follows from the source's own calibration caveat.
Clairic is an extraction model that is able to extract, compute, estimate, and forecast values from unstructured inputs
date — pass: evidence.originalPublished (2026-10-19T17:48:26.614Z) comes from the publisher page and matches the feed entry timestamp. No earlier first publication is indicated.
support — pass: Each claim is backed by an exact source excerpt. The 19.3 s vs 61.7 s figures and the 49% cost reduction match the source, and they are presented as a single company-reported workflow. The caveats on arithmetic chains, aggregation, calibration on unfamiliar data and well-formed but not necessarily correct outputs are all source-stated. No unsupported benefits or impacts are added.
Longer chains of dependent steps are hard for any one-pass model without a scratchpad. The same goes for aggregating across many rows of a table.
media — pass: The selected media is the verified direct link to the original Cloudflare blog post. This is the most direct way for readers to reach the source.
Challenge or correct this reading
Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.