23 Aug 2026 · 1 min · 6 views
FreeToken runs a 753B model on one workstation GPU
FreeToken is a serving engine that runs very large Mixture-of-Experts models on ordinary hardware. It splits the wo...
News
Daily updates on data and AI
8 articles
23 Aug 2026 · 2 min · 8 views
Meta's Pocket lets you build small games by describing them
Meta has brought Pocket, an app that turns prompts into small playable games, to users across the United States.
23 Aug 2026 · 2 min · 7 views
Anthropic puts its most cyber-capable model behind a scanner, not a chat box
Claude Security now runs vulnerability scans powered by Claude Mythos 5, but enterprise teams never talk to the mod...
23 Aug 2026 · 2 min · 5 views
Ramp opens up the AI model router it built for itself
Ramp has released Router, the model-routing service it used internally for three years, free until the end of 2026.
23 Aug 2026 · 1 min · 7 views
Vercel's free tool scores how well your website works for AI agents
Is Agentic runs 118 checks against a public site and grades how easily an AI agent can use it.
23 Aug 2026 · 2 min · 7 views
A 462MB model that cleans up speech-to-text on your own machine
Superwhisper has released S1-mini under Apache 2.0, a small open-weights model that turns raw transcripts into clea...
23 Aug 2026 · 1 min · 8 views
Google ships Gemini 3.7 Flash at half the price of 3.6
Google has released Gemini 3.7 Flash three weeks after 3.6, with better coding scores and half the cost.
8 Jun 2026 · 6 min · 66 views
Diffusion models break out of the image lane: video, language and open weights in 2026
Six months into 2026, the diffusion architecture has expanded well beyond image generation. Alibaba, Inception Labs...