About
All Posts
Partner
Logo
Login
Subscribe
Logo
  • Home
  • Posts
  • GPT-4.5 ‘Orion’ launches in research preview, similar quality to GPT-4o

Feb 28, 2025

GPT-4.5 ‘Orion’ launches in research preview, similar quality to GPT-4o

Bay Area Times
Bay Area Times

Brought to you by:

Top stories today:

  1. GPT-4.5 ‘Orion’ launches in research preview, similar quality to GPT-4o
  2. SEC says most memecoins not securities, “more akin to collectibles”
  3. Meta said to launch dedicated Meta AI app to take on ChatGPT
  4. China vows “all necessary measures” against new U.S. tariffs
  5. Q4 GDP held unrevised at 2.3% while inflation marked higher in Q4

Follow us on X | Instagram | Advertise

0. Data and calendar

All values as of 6 AM ET / 3 AM PT, other than S&P500 and NASDAQ close (4 PM ET / 1 PM PT).

All times are ET.

1. GPT-4.5 ‘Orion’ launches in research preview, similar quality to GPT-4o

GPT-4.5 slightly outperformed GPT-4o, o1, and o3-mini on SimpleQA, which test models on factual questions, with fewer hallucinations, per OpenAI:

OpenAI.

  • Perplexity’s Deep Research outperformed GPT-4.5 on the same benchmark.

  • GPT-4.5 also lagged behind OpenAI’s Deep Research and Claude 3.7 Sonnet on the coding-focused SWE-Bench Verified.

  • “[N]ot a frontier model,” OpenAI initially acknowledged about the GPT-4.5.

  • Initially for ChatGPT Pro, expanding to Plus and Team next week, and then Enterprise and Edu.

  • Also available through Microsoft’s Azure AI Foundry.

Sam Altman has said OpenAI staggered GPT-4.5's rollout because it is “out of GPUs”

X: Sam Altman.

GPT-4.5 is OpenAI’s most expensive model per token, costing $75/M input tokens and $150/M output tokens

X: Artificial Analysis.

  • By comparison, GPT-4o is at $2.5/M input tokens and $10/M output tokens.

2. Temu vs. Amazon: Who’s winning the price war?*

Omnisend

When it comes to online shopping, two names dominate the battlefield — Temu and Amazon. But what really sets them apart? Omnisend dug into the data, analyzing product listings, pricing tactics, and consumer trust on both platforms.

As a result, they’ve compiled a fascinating look into fierce competition, eyebrow-raising discounts, and growing concerns over brand imitation and fake reviews.

Here’s a glimpse of what they found:

  • Temu undercuts Amazon on price: Close product matches are 40% cheaper on average, saving shoppers around $13.37 per item

  • Product overlap is huge: 77% of Amazon-listed products have close equivalents on Temu, and one in 10 are identical

  • Brand imitation raises red flags: From blurred logos to copycat packaging, Temu products often mimic well-known brands

  • Discounts run deep: Temu’s slashing prices with up to 98% off in some categories, compared to Amazon’s 67% max discounts

  • Trust gap is real: Amazon’s review system is robust (50,000 reviews per product on average), while Temu trails behind with just 1,500, plus signs of potential review manipulation

So, who’s really winning — and at what cost?

👉 Get the full breakdown in Omnisend’s latest research.

*Sponsored.

3. SEC says most memecoins not securities, “more akin to collectibles”

  • Thu.: the SEC issued long-awaited guidance on memecoins.

  • The clarification comes after the rapid rise of memecoins following Trump’s comeback.

  • Dogecoin, tied to the Solana blockchain, is among the top memecoins.

  • Memecoins are 3-4x more actively traded than Bitcoin and Ether.

4. Meta said to launch dedicated Meta AI app in Q2 to take on OpenAI’s ChatGPT

Meta.

Subscribe to keep reading

This content is free, but you must be subscribed to Bay Area Times to continue reading.

Already a subscriber?Sign in.Not now

Keep reading

Aug 7, 2026

OpenAI said to plan $300+ hockey puck-sized AI speaker for 2027

OpenAI said to plan $300+ hockey puck-sized AI speaker for 2027

Aug 6, 2026

OpenAI seeks dismissal of Apple’s trade secrets lawsuit

OpenAI seeks dismissal of Apple’s trade secrets lawsuit

Aug 5, 2026

SpaceX Q2 revenue +92% YoY to $7.8B, vs. $6.8B est., net loss narrows to $0.5B

SpaceX Q2 revenue +92% YoY to $7.8B, vs. $6.8B est., net loss narrows to $0.5B

Aug 4, 2026

White House hosts OpenAI, Google, Anthropic on AI oversight framework

White House hosts OpenAI, Google, Anthropic on AI oversight framework

Subscribe

Your morning briefing, free forever.

Get the latest news, resources, and more delivered straight to your inbox everyday.

The visual daily newsletter on business and tech. Analyzing the news with 1 visual per story. Join our 250,000+ subscribers here.

Copyright © 2026 Bay Area Times.

BAT

Archive Posts

Partner

Company

About Us

Legal

Privacy Policy

Terms of Service