ChatGPT just got a new default model. It costs 15x less.
The past two weeks reset the cost floor for AI. OpenAI released GPT-4o mini on July 18, replacing GPT-3.5 Turbo as the default model for free ChatGPT users. At $0.15 per million input tokens, it costs fifteen times less than GPT-4o while scoring 82% on the MMLU benchmark.
What matters today
The past two weeks reset the cost floor for AI. OpenAI released GPT-4o mini on July 18, replacing GPT-3.5 Turbo as the default model for free ChatGPT users. At $0.
Welcome byte
This issue gives executives a clean scan of ChatGPT just got a new default model. It costs 15x less.. Start with the quick hits, then use the companion guides for the workflows worth testing this week.
Quick Hits
Meta: LLaMA 3.1 Released
Meta released LLaMA 3.1 with 405B, 70B, and 8B parameter models. The 405B model rivals GPT-4o on several benchmarks and is fully open source. Read more
Google DeepMind: Math Olympiad AI
AlphaProof and AlphaGeometry 2 solved International Math Olympiad problems at silver medal level. Read more
Runway: Gen-3 Alpha Released
Runway released Gen-3 Alpha with improved motion quality and longer clip generation for AI video. Read more
Apple Intelligence Developer Beta: What's Coming to Your iPhone and Mac
The CEO of a professional services firm has resisted adding AI tools to her iPhone workflow. She uses Claude on her laptop daily. Read more
GPT-4o mini: ChatGPT's New Default Model and Why It Matters for Your Team
The Director of Customer Success at a regional bank ran AI-assisted ticket summarization for three months. With GPT-4o, the cost was manageable for 50 tickets per day. Read more
Top AI Updates
1. GPT-4o mini: ChatGPT's New Default Model and Why It Matters for Your Team
The Director of Customer Success at a regional bank ran AI-assisted ticket summarization for three months. With GPT-4o, the cost was manageable for 50 tickets per day. At 500 tickets per day, the math stopped working. The team went back to manual processing.
2. SearchGPT: OpenAI's Challenge to Perplexity and Google Search
Three workflow patterns where real-time AI search delivers the most executive value
3. Apple Intelligence Developer Beta: What's Coming to Your iPhone and Mac
The CEO of a professional services firm has resisted adding AI tools to her iPhone workflow. She uses Claude on her laptop daily. The hesitation is practical: a separate app, a separate login, and a separate subscription for mobile AI does not fit how she works between meetings.
Pro Tip
Three specific workflows where GPT-4o mini economics change the build-vs-manual decision
Action items:
- Copy the prompt or framework into ChatGPT, Claude, or Gemini.
- Test Using GPT-4o mini for High-Volume Document Processing and Classification on one live decision, document, or meeting artifact.
- Save the best version as a reusable team template.
Copy-ready guide Using GPT-4o mini for High-Volume Document Processing and Classification
Productivity Gem
A three-step conversational research workflow for competitive intelligence using AI search
Action items:
- Choose one recurring workflow with a clear before and after.
- Run the Building a SearchGPT Research Workflow for Competitive Intelligence workflow once with real work, not sample data.
- Keep it only if it saves at least 30 minutes this week.
Put it to work Building a SearchGPT Research Workflow for Competitive Intelligence
Health Tip
Optimize your executive performance by turning raw biometric data into a personalized recovery plan.
This information is for educational purposes only and is not medical advice. Please consult a qualified healthcare professional before making changes to your health routine.
Step-by-step guide Build a Weekly AI Energy Check From Apple Health
Kids Tip
Kids create a hands-on AI project: Build a Teachable Machine Image Classifier With Your Kid. A parent or educator helps them build, test, and explain what the AI tool gets right and wrong.
Try the activity Build a Teachable Machine Image Classifier With Your Kid
Wrap Up
Pick one workflow from this issue, run it with real work, and keep the result only if it makes the next decision clearer or faster.
About the author
Pierre has spent 25+ years building growth systems across fintech, real estate, lending, campaigns, and AI workflows, with $1.5B+ in client value delivered.
Email usThree deep dives. Four useful moves. One email worth opening.
PromptHacker turns the AI firehose into practical next steps for work, health, family, and everything time keeps trying to steal.