Anthropic launches Claude Haiku 4.5, AI performance, cutting cost and latency
We just got word that Anthropic released Claude Haiku 4.5. It’s a smaller model, yet the company says it holds its own against the older flagship versions.
Latest breakthroughs in large language models and generative AI shaping the future of artificial intelligence and machine learning.
We just got word that Anthropic released Claude Haiku 4.5. It’s a smaller model, yet the company says it holds its own against the older flagship versions.
It looks like Google’s AI Pro plan, usually ₹1,950 a month, is being offered for just ₹11 per month, for three months. I saw the price on the India pricing page, the same spot where AIM first flagged it.
On Wednesday Anthropic launched Claude Haiku 4.5, pitching it as a slimmer cousin of its flagship. The company calls it a “small” language model, yet internal tests say it hits the same performance levels Claude Sonnet 4 hit back in May.
Anthropic has done something that catches the eye: it put Claude Haiku 4.5 on the Claude.ai site for anyone who signs up for free. In other words, the model that the company calls powerful is now reachable without a price tag.
On October 13, 2025 Andrej Karpathy pushed a new repo called nanochat. It’s a PyTorch-based language model that squeezes most of its logic into about 8,000 lines.
Veo 3.1 just landed, and it adds a handful of tweaks to the Flow AI filmmaking suite.
OpenAI seems set to lock ChatGPT behind an 18-plus gate, according to a quick note from The Rundown AI. The headline “ChatGPT to go 18+” hints at a pretty big policy tweak.
Thinking about Google’s cloud storage? The Google One Premium Plan lands right in the middle of the options, costing $10 a month or $100 if you pay yearly.
When Salesforce and OpenAI announced their latest move on Tuesday, it felt like they were finally putting the pieces together.
I keep catching developers with a ChatGPT, Claude, or Gemini tab open, like a coworker who never sleeps. They pull it up for quick boilerplate snippets or to untangle a stubborn bug.
OpenAI just announced a new safety and security committee, but the lineup is already sparking debate. The company says the group - called a “wellness” council - will steer its safety strategy, especially for younger users.
It looks like ChatGPT is about to get a lot more personal. On Tuesday, OpenAI’s CEO Sam Altman said the bot will soon allow “erotica” for users who can prove they’re over 18, tied to an age-gating feature slated for a December rollout.
Learn to build AI-powered apps without coding. Our comprehensive review of No Code MBA's course.
Curated collection of AI tools, courses, and frameworks to accelerate your AI journey.
Get the week's most important AI news delivered to your inbox every week.