Research library
Insights
Analysis of AI, automation, and software: what is changing, what holds up, and what it means for practical decisions.
198 articles · Page 2 of 17
GitHub Actions Lockfiles Pin Every Dependency to a Commit. They Don't Vet Any of Them.
GitHub's Actions lockfile preview pins actions, transitive deps and reusable workflows to commits. What it fixes, what it leaves to you, and what to test.
Liquid LFM2.5-24B-A2B: A Laptop-Sized Agent Model With a $10M Revenue Line
Liquid's LFM2.5-24B-A2B runs 2.45B active parameters and leads its own tool-use tables. It is free for commercial use only below $10M annual revenue.
NVIDIA RTX Spark and Nemotron 3.5 Bolt: A 74B Model in 24GB, Before Anyone Can Test It
NVIDIA's Bolt packs a 74B-A9B MoE into 24GB with 2-bit QAT; RTX Spark offers up to 128GB. What fits, what's unpublished, what to test first.
Falcon ASR Targets Arabic and Emirati Transcription
TII's Falcon ASR focuses on Arabic and Emirati speech. Read the benchmark scope, demo availability and practical evaluation limits.
LightOnOCR-3 Adds Layout and Chart Extraction
LightOn's three OCR models add document regions, image descriptions and chart data. How to evaluate the output before using it in retrieval.
Meta Releases Muse Wonder and Muse Wonder Full
Meta's Muse Wonder models target sustained agent work. What is available now, what Active Memory means, and what teams should test.
Anthropic's Cyber Verification Program Now Has Three Tiers. Which One to Apply For
Anthropic split its Cyber Verification Program into Defense, Red Team and Specialized tiers. Who qualifies, review times, the retention catch and Bedrock.
Claude Haiku 5.5: A 90% List-Price Cut, a 75% Real One, and Five Ways to Get a 400
Haiku 5.5 lists at a tenth of Haiku 4.5's price; Anthropic says about 75% less in practice. The gap, the 400s after switching, and Sonnet 5.5 vs Luna.
Gemini API Preview Shutdowns: Veo 3.1 Moves to Omni, Voice Follows in November
Veo 3.1 and Omni previews may end from October 22, TTS and Live previews from November 17. Cost per clip, voice replacements and the January price rise.
Mistral Large 4: What Is Confirmed Before the Weights Drop, and What Is Not
Mistral Large 4 is a 1T-parameter MoE you can call today at a half-price preview. Weights, license and independent scores are still due. What to check.
Decision Models Go Open Weight: Cloudflare Clef, Liquid d1 and TypeSafe Jev Compared
Cloudflare's Clef, Liquid's d1-3B, vLLM-SR's d3 and Perplexity's pplx-decider: open-weight decision models compared with Jev and OpenAI's Decisions API.
OpenAI Shuts Down GPT-4, o1 and o4-mini on October 23; Copilot Drops GPT-5.5 on October 19
OpenAI's API removes gpt-4, o1, o3-mini, o4-mini and gpt-image-1 on October 23. Copilot removes GPT-5.5 and GPT-5.4 on October 19. A dated checklist.