Short on time? Here’s today in three lines:

  • Anthropic is now a public company. It priced its IPO at $83 a share and started trading on Nasdaq as ANT, in the largest IPO by money raised on record.
  • Google launched Gemini 4 Neon, a mid-tier model it says matches frontier rivals on key tests for far less per task.
  • Superhuman made Go its main AI agent across Mail and Docs, with 20 partner agents plugged in.

1. Anthropic goes public in a record IPO

Anthropic priced its offering at $83 per share and began trading on the Nasdaq Global Select Market under the ticker ANT on 19 November. It sold 980 million Class A shares, including about 23.6 million from existing holders. Fortune reports the price came in above the $70-$78 range, values the company at about $1.5 trillion and raises about $79 billion for Anthropic itself, ahead of SpaceX’s $75 billion in June.

Not everyone is cheering. TechCrunch notes that Amazon and Google are both top shareholders and top suppliers, so much of their investment comes back as compute spending. Anthropic reported $17 billion in third-quarter revenue but still runs a large net loss. At the opening bell, CEO Dario Amodei repeated his pledge to pace AI development until the company is confident its systems are safe.

Why it matters: Public companies must report results every quarter. Buyers of Claude will now see far more detail on revenue, costs and risks than any private AI lab shares.

Our take: This is good news for procurement teams. A listed vendor with audited numbers is easier to assess than a private one. But quarterly pressure also pushes prices and product roadmaps. Lock in multi-year terms where you can, and keep a second model provider tested and ready. If you use Claude today, start with our Claude review.

2. Gemini 4 Neon: near-frontier smarts at a mid-tier price

Google released Gemini 4 Neon on 18 November. It sits between Hydrogen and Carbon in the Gemini 4 family and replaces Argon. AI Pro and Ultra subscribers can pick it as “Deeper analysis” in the Gemini app. It has a 1 million-token context window and four thinking levels.

On price, Digital Trends reports API rates of $3 per million input tokens and $15 per million output tokens. Google claims Neon scores 61 on the Artificial Analysis Intelligence Index, ahead of Claude Opus 5.5 (59) and GPT-6.2 Sol (57). Google also says it costs 50-80% less per task than leading frontier models. These are Google’s own figures, and Neon still trails rivals on some coding and chart tests.

Why it matters: Most business AI work is routine: summaries, emails, data pulls. A cheaper model that is nearly as smart cuts the cost of running agents at scale.

Our take: Don’t switch on a benchmark chart. Run your own top 20 prompts through Neon and your current model, then compare quality and cost per task. Watch output tokens: they usually drive most of the bill. See our Gemini review for plan details.

3. Superhuman wants Go to be your one work agent

Superhuman, the company formerly called Grammarly, is making Go its central agent across Superhuman Mail and Docs. Go works inline as you type, learns your writing style and can act on its own: checking schedules, watching email and drafting replies. In Mail it replaces the old AI tools and adds search across company knowledge and connected apps.

Through a new SDK, 20 partner agents plug into Go, including DeepL, Amplitude and Databricks. Superhuman is also rolling out shared live meeting notes after buying Fathom in September. CEO Shishir Mehrotra says Go has more than 1 million active users.

Why it matters: The fight is shifting from chatbots to agents that sit inside the tools you already use. Whoever owns your inbox and documents owns your daily workflow.

Our take: Inbox agents save real time on follow-ups and scheduling. But an agent that drafts and acts across Mail, Docs and 20 partner apps also sees everything. Before rollout, check which data Go can reach and who can turn on partner agents. Compare it with our Superhuman review.

Quick hits

What this means for travel and hospitality

Hotels and travel firms buy AI mostly through software vendors, and those vendors run on models from Anthropic, Google and OpenAI. Anthropic’s listing and Neon’s pricing both point the same way: capable models are getting cheaper per task, and the big providers now face more pressure to show steady revenue. Expect vendors to pass some savings through, and to push more agent features. Guest messaging, review replies and rate-shopping summaries are prime candidates. Canva’s off switch is a useful reminder too: brand teams want control over when AI touches their creative work.

One move for Monday

List every AI tool your team pays for and note which model sits underneath each one. Ask your top two vendors whether they can switch to a cheaper model, such as Gemini 4 Neon, and what that would save you per month. Then check how your brand shows up in AI answers with our free AI visibility audit.

Get the week’s biggest AI moves in one email with the Digicrusader weekly brief, or catch up on every update in AI News.

Anurag Jain

Anurag Jain

Founder, Digicrusader

E-commerce, digital marketing and AI leader with experience at Amazon, Expedia, Kenvue, Godrej and RateGain. Author of Powerhouse: An AI Playbook.