Blog
Jun 25, 2026 · 20 min read
OCR’ing 100,000 pages with open-source VLMs on Modal
We OCR'd 100,000 pages with open-source vision-language models in <1hr for $223, roughly 9–27× cheaper than the comparable-quality proprietary APIs.
Read more → Jul 15, 2025 · 8 min read
We Benchmarked Frontier Reasoning Models on The Atlantic's Bracket City
A deep dive into how frontier AI models perform on complex word puzzles, revealing surprising insights about reasoning efficiency vs. accuracy.
Read more →Join Our Newsletter
Get a summary of what we've shipped during the last month and behind the scenes updates, and team picks.