What are the LLMs with the most output tokens these days? GPT-4 and Claude 3 are both 4096. Gemini Pro 1.5 is 8192 This really matters for structured data extraction: even with 1m of input tokens you can't scrape a big webpage into a CSV file if you run out of output tokens
AI
-

DataAI Summit 2026: 500+ Sessions on Data, Governance, GenAI
By
–
With over 500+ sessions covering everything from data warehousing to governance to the latest in #GenAI, #DataAISummit has something for everyone! Check out the upcoming sessions and save your spot https://
bit.ly/3UlaDA7 -
Claude 3 Opus Vision Models Refuse PII in PDF Analysis
By
–
I have been doing a bunch of experiments with the latest vision models and PDFs recently, they're pretty great! Claude 3 Opis did point-blank refuse to answer questions about one that included personally identifiable information though
-
Effective Chunking Strategies for GPT-3.5 Implementation
By
–
Have you found a chunking strategy that works well powered by gpt-3.5? I am hoping to add chunking to LLM at some point but I'm not sure what strategies I should include
-
Browser-based packages loadable from CDN networks
By
–
Do you know anything like that which runs directly in the browser via a package you can load from a CDN?
-
Expanding Tesseract.js Language Support in OCR Tool
By
–
Tesseract.js works with other languages, but I've not exposed that in my tool yet – I should do that!
-

Seven Free Online AI Courses Available from NVIDIA
By
–
RT @AndrewBolis 7 FREE online courses in #AI from @nvidia @NVIDIAAI
-
textract-cli: Lightweight AWS Textract Command Line Wrapper
By
–
My other OCR project from yesterday: textract-cli, a tiny CLI wrapper around AWS's amazing but so-hard-to-use Textract API https://
github.com/simonw/textrac
t-cli
… Assuming you have AWS credentials configured: pipx install textract-cli
textract-cli image.jpeg > output.txt <5MB JPEG/PNGs only -
Signal addresses security bug reports through official channels
By
–
The ‘actual user’ has not shared debug logs or reached out to us either at support@signal.org or security@signal.org—two widely publicized channels—before blasting conjecture on Twitter. Meanwhile we’ve seen this behavior in other reports, looked into it & are working on a fix.