A user reports that a search assistant configured with the 'academic' lens failed to restrict its searches when running on Gemini 3.1 Flash, returning non-academic websites instead, while the same fil…
A Hacker News discussion flagged stack-junkie.com as AI slop appearing in Kagi's Small Web results, with user fxgn noting the site was added to the smallweb.txt list in a GitHub commit on August 6 and…
A new open-source tool, Evalshift, provides a repeatable method for testing large language model (LLM) migrations by freezing a golden suite of recorded agent runs, running both source and target mode…
XAI launched Grok Voice Think Fast 2.0, a speech-to-speech model that scored 82.9% on Artificial Analysis' benchmark, up from 75.7% for version 1.0 and ahead of GPT-Realtime-2.1 at 79.1%. The model co…
LLM API costs have dropped 94.5% since April 2023, with Gemini 3.1 Flash now costing $0.40 per million output tokens compared to GPT-4's $60 in March 2023, according to BenchLM.ai's pricing index. Mos…
Google released Nano Banana 2 Lite, a faster and cheaper AI image generator, on Tuesday. The model produces images in 4 seconds at $0.034 per 1,000 images, targeting high-volume workflows. The launch …
Google released multiple Workspace updates on June 17, 2026, integrating Gemini with Google Classroom for education, enabling longer Veo videos and parallel generation in Google Vids, enhancing AI ava…
FLUX.2 and Gemini 3.1 Flash produced the strongest character consistency across three tests comparing OpenAI, Gemini, Black Forest Labs, and Runway image generators. gpt-image-2 placed third and Runwa…