1.
3.
F1 in Madrid: Like Monaco but twice as long and none of the glamour
(arstechnica.com)
4.
Kimi-maker Moonshot AI targets $2B in annual revenue
(techcrunch.com)
5.
Kimi-maker Moonshot AI targets $2 billion in annual revenue
(techcrunch.com)
6.
Moonshot serves Claude instead of Kimi and collects exchanges for model training
(news.ycombinator.com)
7.
Cognition's SWE-2 achieves 92.8 on Terminal-Bench 2.1
(news.ycombinator.com)
8.
Cognition launches new SWE-2 model, Rivaling Fable 5.1 and GPT-Astra
(news.ycombinator.com)
9.
Harvey hits $15.5B valuation, months after reaching $11B
(techcrunch.com)
10.
Qwen 3.8 follows GPT-5.5 Pro reasoning prefills
(news.ycombinator.com)
11.
Kimi K3 (2.8T) at 1 token/s on a MacBook Pro, streamed from four SSDs
(news.ycombinator.com)
12.
Compute-efficient pretraining and scaling to trillion-parameter models
(news.ycombinator.com)
13.
vLLM v0.28.0
(news.ycombinator.com)
14.
Kimi Work attaches raw agent sessions to feedback reports
(news.ycombinator.com)
15.
Choosing an AI model: one prompt, 11 models, different results
(news.ycombinator.com)
16.
17.
A New Trick Reveals AI Models’ Inner Thoughts
(wired.com)
18.
20.
Smaller, faster, safer: running Kimi and GLM at scale
(news.ycombinator.com)
21.
AirLLM 70B inference with single 4GB GPU
(news.ycombinator.com)
22.
Running Kimi K3 on MI355X at Better Performance per Dollar Than B300
(news.ycombinator.com)
23.
24.
Run Kimi K3 using 29 GB of RAM at 0.50 tok/s
(news.ycombinator.com)
25.
Moonshot’s Kimi uses 20k Nvidia chip cluster from Alibaba
(news.ycombinator.com)
28.
Kimi K3-256k
(news.ycombinator.com)
29.
30.
Running Kimi K3 on a M1 Max
(news.ycombinator.com)
Today's top topics:
android authority
artificial intelligence
anthropic
donald trump
openai
polymarket