Cloudflare Introduces Clairic, Its Inaugural System One Extraction Model – Unite.AI

Cloudflare Unveils Revolutionary Extraction Model Clairic

On October 19, 2026, Cloudflare introduced Clairic, a groundbreaking System One extraction model that promises enhanced accuracy, speed, and cost-efficiency. This innovative model is hosted on Workers AI, with plans for an open-weights release in the near future.

Availability and Pricing on Workers AI

Clairic, available as @cf/cloudflare/clairic, can be accessed through a Cloudflare Worker using env.AI.run() or the REST API. The model documentation highlights a 65,536-token context window, vision support, and a unit pricing of $0.10 per million input tokens.

How the Extraction Model Works

Clairic functions by extracting typed values from unstructured inputs, providing accurate responses in under a second. Customer examples demonstrate the model’s ability to compute, estimate, and forecast with precision.

Documented Examples and Reported Tests

Cloudflare showcases real-world scenarios where Clairic excels in computing totals, extracting information from receipts, and providing accurate estimates and forecasts. The model’s performance in various tests demonstrates its reliability and efficiency.

Preview Limitations and Open-Weights Plans

While Clairic shows great promise, Cloudflare acknowledges some limitations in its current state. However, the company plans to continue refining the model and releasing open-weights to enhance its capabilities.

  1. What is Cloudflare Releases Clairic?
    Cloudflare Releases Clairic is the company’s first system one extraction model, designed to extract information from unstructured text data.

  2. How does Clairic work?
    Clairic uses natural language processing (NLP) techniques to analyze and extract relevant information from text data, allowing users to gain insights and make data-driven decisions.

  3. What are the potential applications of Clairic?
    Clairic can be used in a variety of industries and applications, such as sentiment analysis, document classification, and content analysis.

  4. How accurate is Clairic in extracting information?
    Cloudflare has tested Clairic extensively to ensure high accuracy in extracting information from text data, and continues to refine and improve the model through ongoing research and development.

  5. Can Clairic be integrated with existing systems and platforms?
    Yes, Clairic is designed to be easily integrated with existing systems and platforms, making it a versatile and powerful tool for businesses looking to unlock the value of their unstructured text data.

Source link

Laude Institute Unveils Inaugural Cohort of ‘Slingshots’ AI Grants

Laude Institute Launches Innovative Slingshots Grants to Propel AI Research

On Thursday, the Laude Institute unveiled its inaugural Slingshots grants, focused on enhancing the development and application of artificial intelligence.

Empowering Researchers with Essential Resources

The Slingshots program serves as an accelerator for researchers, offering vital resources often lacking in traditional academic settings. These include funding, computational power, and product engineering support. In return, grant recipients commit to delivering tangible outcomes—be it a startup, an open-source codebase, or another noteworthy creation.

First Cohort Tackles AI Evaluation Challenges

The program’s initial cohort comprises 15 projects, primarily targeting the intricate issue of AI evaluation. Among the featured initiatives are well-known projects such as Terminal Bench, a command-line coding benchmark, and the evolving ARC-AGI project.

Innovative Solutions from New Projects

Some projects introduce novel strategies to long-standing evaluation issues. For instance, Formula Code, developed by Caltech and UT Austin researchers, aims to assess AI agents’ capacity for optimizing existing code. Meanwhile, BizBench from Columbia proposes a comprehensive benchmark for evaluating “white-collar AI agents.” Additional grants are dedicated to exploring new frameworks for reinforcement learning and model compression.

Dynamic Competition: The CodeClash Initiative

John Boda Yang, co-founder of SWE-Bench, is leading a new initiative called CodeClash as part of this cohort. Drawing inspiration from SWE-Bench’s success, CodeClash aims to evaluate code through an engaging, competition-focused framework.

Insights from Industry Experts

“I believe that ongoing evaluations based on core third-party benchmarks drive progress,” Yang shared with TechCrunch. “My concern is the potential future where benchmarks become too tailored to specific companies.”

Here are five FAQs regarding the Laude Institute’s announcement of the first batch of "Slingshots" AI grants:

FAQ 1: What are the "Slingshots" AI grants?

Answer: The "Slingshots" AI grants are funding opportunities offered by the Laude Institute aimed at supporting innovative projects that leverage artificial intelligence. These grants are designed to promote groundbreaking research and development in the AI sector.

FAQ 2: Who is eligible to apply for these grants?

Answer: Eligibility for the "Slingshots" AI grants typically includes researchers, academics, startups, and organizations focused on AI initiatives. Specific eligibility criteria may vary, so it’s essential for potential applicants to review the guidelines provided by the Laude Institute.

FAQ 3: How much funding is available through the "Slingshots" AI grants?

Answer: While the exact amount of funding may vary by project, the "Slingshots" AI grants offer significant financial support to selected projects. Interested applicants can find detailed information on the funding range in the grant guidelines available on the Laude Institute’s website.

FAQ 4: What types of projects are prioritized for funding?

Answer: The "Slingshots" AI grants prioritize projects that demonstrate innovative uses of artificial intelligence, including but not limited to machine learning applications, AI in healthcare, environmental sustainability, and solutions addressing societal challenges. Projects that align with these themes are encouraged to apply.

FAQ 5: How can interested applicants apply for the grants?

Answer: Interested applicants can apply for the "Slingshots" AI grants by visiting the Laude Institute’s official website. There, they will find the application form and guidelines, including deadlines and submission requirements. It’s recommended to prepare a comprehensive proposal outlining the project’s goals, methodology, and expected impact.

Source link