Monday, May 4, 2026

xAI’s New Model Push

xAI’s New Model Push

Today’s Overview

Good morning, xAI is pushing Grok 4.3 with lower pricing and a broader product plan, while Anthropic keeps drawing attention from both investors and enterprise partners. Meanwhile, new research says an older OpenAI model outperformed doctors in an ER study, and Maryland just drew a hard line on AI grocery pricing. Let's dive in.

Top Stories

xAI Launches Grok 4.3

xAI launched Grok 4.3 as a new reasoning model aimed at coding, logic, math, scientific work, and agentic tool use. The release also comes with lower pricing than Grok 4.20 and a wider push into voice and audio features. The model update lands alongside fresh attention on xAI's training practices and the broader competition around frontier models.

  • Artificial Analysis says Grok 4.3 improved xAI's Intelligence Index score while cutting input and output pricing versus Grok 4.20.
  • The product roadmap also includes Custom Voices and the Voice Library, plus the Imagine API for native audio generation for video.
  • xAI also says Grok Voice mode is coming to Apple CarPlay as the company expands where the model can be used.

Maryland Bans AI Grocery Pricing

Maryland signed the U.S.'s first ban on AI-driven grocery pricing. The law targets the use of personalized shopper data to mark up prices and includes fines of up to $25,000 for stores caught doing it. It is one of the clearest signs yet that regulators are moving from AI hype to AI restrictions.

  • The new law is the U.S.'s first ban on AI-driven grocery pricing.
  • It targets stores using personalized shopper data to adjust what customers pay.
  • Violations can bring fines of up to $25,000.

Anthropic Nears Huge New Round

Anthropic reportedly moved to close a roughly $50 billion round that could value the company around $900 billion or higher. The reported deal reflects strong investor demand and rapid revenue growth nearing a $40 billion run rate. It underscores how aggressively capital keeps chasing the biggest AI labs.

  • The company is reportedly closing a roughly $50 billion round.
  • That could value Anthropic at around $900 billion or higher.
  • The reported growth story includes revenue nearing a $40 billion run rate.

Research & Analysis

Old AI Model Beats Doctors In ER Trial

A Harvard study published in Science put OpenAI's o1-preview through 76 real emergency room cases across three stages of patient care. At initial triage, the model gave the correct diagnosis more often than two attending physicians. Two physician reviewers could not reliably tell which diagnoses came from the model and which came from humans.

  • At initial ER triage, the model was correct 67.1% of the time.
  • The two physicians scored lower at 55.3% and 50.0%.
  • In one case, the AI flagged a rare flesh-eating infection 12 to 24 hours early.

Trending AI Tools

  • Codex Adds animated Pets overlays, config imports from other coding agents, and a dictation dictionary.

  • Manex Preserve useful answers, corrections, and context as memory.

  • Aaavatar Branded team headshots in one drop.

Quick Hits

Keep reading for free

Enter your email. If you're already subscribed, we'll send a sign-in code. If not, you'll subscribe in the next step.

Free access. Subscribe once, then use the same email on future issues.

Free to read. Subscription just unlocks the full issue.