Quilty integrates script analysis with Claude and ChatGPT, and allows for comparison of scores between old and new drafts.
PR Newswire
1h ago
Ai Focus
Quilty announces that its AI script analysis platform has been integrated with Claude and ChatGPT. Users can now directly view the coverage analysis and compare the score changes between old and new drafts within the assistant.
Helpful
No.Help

Directors and screenwriters can now directly process the overlay analysis notes within their existing assistants, and then let Quilty determine whether the new drafts have indeed improved.

Los Angeles, October 8th / PRNewswire / -- Quilty Co-founder, Emmy, Tony, and Olivier Award-winning producer Simon Horsman, as well as tech entrepreneur Daniel Wood, announced today that this entertainment intelligence platform AI for script analysis is now available on Claude and ChatGPT. Screenwriters can open a chat window, connect to their own Quilty accounts, and directly use the content they already have, such as coverage analyses, story notes, character work, market and talent insights, without having to export PDF or paste reports into the window.

In addition to these integrations, Quilty also highlights a feature that is directly aimed at revising manuscripts: draft-by-draft comparison. After the screenwriter submits a new draft to Quilty, they can ask the assistant to compare it with the previous version and provide a report on the changes made in various dimensions back to Quilty Score.

Quilty CEO Horsman stated: "Integrating analysis into Claude and ChatGPT means that writers can take a note, try out a new scenario, and then return to Quilty to see if this effort has truly led to a change in scores. This is developmental work, not just a file sitting in a download folder."

The design goal of Quilty Analysis is to function like a comprehensive analysis tool for readers, rather than simply dumping data. A complete functional analysis using Quilty goes through 33 stages and generates a report that encompasses over 70 different sections, covering story creation, writing techniques, characters and dialogue, market and finance, talent packaging, audience and culture, production, and more. This integration is intentionally designed not to overwhelm users with all the content at once.

13 read-only tools allow assistants to start where professional readers begin – Quilty Score – as well as a limited briefing that outlines the advantages, issues, and priority actions; then they can delve into a specific chapter, section, or particular scenario. Screenwriters can perform semantic searches on the main text of their scripts, search for analysis reports, retrieve talent packaging or comparative film studies, or apply for a link to access the complete PDF for a short period of time.

Wood, Quilty Chief Technology Officer stated: "The complete report is still there when you need a printable version. But Claude shouldn't try to digest 70 chapters at once with the first question. First, look at the scores and main points, then delve deeper—this is how we aim to cover the needs of analysis readers, and it's also how we build our servers."

The foundation of these integrations is Quilty Score V5, which is a comprehensive score ranging from 0 to 100. It consists of four weighted pillars: story and technique, commercial feasibility, cultural resonance, and production reality, and is further broken down into 15 sub-dimensions. The score is generated by a cluster of models and presented with confidence intervals and model consistency indicators, rather than a single number that appears definitive.

By default, the scoring process is blind: before the script is fed into the scoring model, identifying information is removed, so a well-known work will be judged based on its content rather than its fame. Quilty Score is a diagnostic assessment of the submitted material, not a box office prediction.

What supports this engine is a system that deliberately adopts a multi-supplier architecture. Quilty routes each stage to the most suitable model for that stage. Using Claude from Anthropic, GPT from OpenAI, Grok from xAI, Gemini from Google, Perplexity and DeepSeek – creativity and skill judgment are entrusted to one type of model, cultural and financial reasoning to another type of model, and real-time market research to the search native model. There is no blind spot of any single supplier that will become a blind spot for Quilty.

For more information, see:

https :// drive.google.com / drive / folders /13oBHVh16vczU- Sy7yhA6JMqo6fqKwgun ? usp = sharing

  • Working principle, explained step by step.
  • Insights from user Quilty, screenwriter/director/producer Nick Lyon.
  • Press releases Word and PDF versions

About Quilty

Integration Documentation: quilty.app / docs / claude-connector

Claude Desktop and Claude Code can be connected through the personal token in Settings.

Connectors on quilty.app. Same server, same range: https :// quilty.app / api / mcp.

Contact: David Roberson

Roberson Public Relations

[ email protected ]

310-925-8786

Tip
$0
Like
0
Save
0
Views 20
CoinMeta reminds readers to view blockchain rationally, stay aware of risks, and beware of virtual token issuance and speculation. All content on this site represents market information or related viewpoints only and does not constitute any form of investment advice. If you find sensitive content, please click“Report”,and we will handle it promptly。
Submit
Comment 0
Hot
Latest
No comments yet. Be the first!
Related
Gasoline prices are unlikely to fall significantly before the election day, according to market forecasts
Predictive market platform Kalshi shows that traders believe that by the election date of November 3rd, the national average gasoline price in the United States AAA still has an 87% probability of remaining above $4 per gallon; diesel prices are also expected to remain high. The report indicates that the persistently high fuel prices are affecting the election outcome, with the Democratic Party's chances of regaining control of the Senate rising to Kalshi.
CNBC
·2026-10-09 03:00:42
11
Anthropic Updates Usage Policy to Prohibit Model Abuse and Interference in Elections
Anthropic Policy Update on Thursday: New Prohibitions on Electoral Interference, Weaponized Software, and Monitoring; Also, Continuous Verbal Abuse of Models by Users is Strictly Forbidden. The company stated that these regulations only apply in extreme circumstances and are not intended for common user dissatisfaction, rebuttals, dark-themed creativity, or model testing and research.
TechCrunch
·2026-10-09 03:00:40
8
Lemonade Adds support for Tesla FSD V14 Lite, extending autonomous vehicle insurance to Teslas equipped with HW3
Lemonade announces that its Autonomous Car insurance now covers more Tesla models. Owners in Arizona, Colorado, and Tennessee who are equipped with HW3 and use the new FSD ( Supervised ) v14 Lite can save 30% on the mileage covered by FSD; owners who are equipped with HW4 and use FSD v14.2 or an updated version can still save 50%.
PR Newswire
·2026-10-09 03:00:39
11
Reportedly, OpenAI's revenue is 20 billion US dollars lower than previously expected.
According to the Financial Times, OpenAI has stated to investors that its annual revenue is "nearly $50 billion," which is about $20 billion lower than the previously circulated figure of around $70 billion. The report also mentions that the figure of $70 billion comes from information previously shared by OpenAI with investors and was used for direct comparison with the annual revenue of Anthropic.
TechCrunch
·2026-10-09 02:51:15
15
San Diego jury awards $2 million in compensation in genital herpes transmission case
According to SBEMP LLP, a San Diego jury awarded Tiffany Young $2 million in compensation in a lawsuit against the former CEO of Funko, Brian Mariotti. The jury found Mariotti responsible for making intentional false statements, and the plaintiff claimed that he transmitted genital herpes to her during their intimate relationship.
PR Newswire
·2026-10-09 02:51:12
14
View More