All times UTC

The AI timeline

Launches, drama, outages, weird experiments and the moments changing today's conversation.

800 events
13 updates
37 updates
Bots in Production Ecosystem Community reports

17 bots for $100 a month? Only three call a model

An operator's newly shared cost breakdown puts 17 automation jobs on one Claude subscription, but fourteen are ordinary Python and the bill excludes labor and existing equipment.

3 sources Open story
Qwen Image Turbo Multi-format Quantizations Momentum Community reports

A 4.05GB Qwen Image Turbo GGUF has a bigger setup bill

abenzerps's multi-format package appears in a fresh Hugging Face trending snapshot, with GGUF, Safetensors and MLX options—and separate companion models that matter for setup.

3 sources Open story
39 updates
32 updates
Brave publisher content-scraping lawsuits Debates Developing

Publishers take the AI scraping fight to Brave

The New York Times and other publishers have sued Brave, TechCrunch reports, extending the copyright fight to a company alleged to supply content to AI services.

1 source Open story
Moebius Inpaint GGUF Launches Community reports

Moebius image repair gets a 247MB community GGUF

cstr has published four image-inpainting variants for CrispEmbed, including a mixed-precision Q4 file. The downloads need a specific experimental engine and a separate VAE.

6 sources Open story
Keural-Lab-1031 Launches Community reports

Keural's Qwen experiment lands as a 78.8GB GGUF

The experimental community checkpoint now has a Q4_K_M download and optional vision projector. Its author's scorecard trails the comparison model on all five reported benchmarks.

6 sources Open story
Krea 2 NVFP4 Matmul Patch Ecosystem Community reports

Krea 2 NVFP4 patch claims a 2.5x speedup on RTX 5080

A community Python script clears a metadata flag in ComfyUI's Krea 2 Turbo NVFP4 file. Its author reports 11.6-second warm renders falling to 4.6 seconds, with quality tradeoffs still open.

4 sources Open story
14 updates
9 updates
30 updates
32 updates
27 updates
19 updates
36 updates
12 updates
33 updates
27 updates
31 updates
37 updates
Qwen Code Ecosystem Community reports

Qwen Code 0.25.2 lets Mods intercept agent tool calls

Opt-in experimental Mods can rewrite tool arguments, block calls and inspect permission decisions in trusted headless main sessions. TypeScript support and selected Claude input contracts arrive with clear limits.

4 sources Open story
40 updates
Ternary Bonsai 2 27B NInfer-3090 Launches Community reports

Bonsai 2's RTX 3090 conversion reports 297 tokens/s

A new 8.87 GiB native NInfer artifact brings Prism ML's existing ternary 27B model to the RTX 3090 fork. Its author reports 297 tokens/s on a short coding prompt with DFlash2.

4 sources Open story
35 updates
36 updates
25 updates
6 updates
2 updates
11 updates
27 updates
30 updates
33 updates
20 updates
10 updates
8 updates
6 updates
29 updates
19 updates
12 updates
OpenAI dots Industry Developing

OpenAI dots stalls during its first public demo

Futurism reports that Dottie acknowledged a request about user testing but did not deliver the requested update during the live demo. The technical cause remains unestablished.

Live-demo interruption 2 sources Open story
McDonald's AI pricing engine Debates Developing

Reuters details McDonald's AI-assisted menu pricing engine

Reuters found a machine-learning system recommending restaurant-level menu prices using transaction data and estimates of what local customers will pay. McDonald's calls it an optional tool.

Nearly 14,000 US restaurants 1 source Open story
5 updates
12 updates
6 updates
Meta Muse Industry Source-backed

Amazon blocked Meta's Muse shopping agent

The hottest consumer-agent launch immediately ran into the question every agent company eventually faces: who gets permission to act on someone else's platform?

146 HN points 3 sources Open story