Past 30 days of Claude Code usage on Claude Max $100 plan = ~4.13 billion tokens Opus 5 family = 46% Opus 4.8 family = 20% Fable 5 family = 22% Sonnet 5 family = 6
George
@georgesl.bsky.social
My AI Journey Blog https://ai.georgeliu.com Centminmod.com LEMP stack maintainer. Cloudflare Community MVP since 2018
Soulm8te one of our possible futures with AI and robotics www.youtube.com/watch?v=_WwB... ?
SOULM8TE Official Trailer | Blumhouse AI Horror Thriller
YouTube video by FlickDirect: Trailer Vault | Ultimate Trailer Fix
youtube.com
More folks using ChatGPT replacing lawyers in court. While I use AI, definitely not ready for that level of trust! Would you trust AI representing you in court? youtu.be/gimPnqb5ubU?...
youtu.be
Stretching my Claude Max $100 plan with newly upgraded ChatGPT Pro $100 plan using Claude Code /codex-delegate skill so Claude tells GPT-5.6 to do the work 🤓
First 24hrs after upgrading to ChatGPT Pro $100, used 35% of my weekly usage limits!
Upgraded to ChatGPT Pro $100 within ChatGPT desktop app to join my Claude Max $100 plan and other subscriptions 🤓
Turbo Fieldfare runs a 26 billion parameter Gemma 4 model on a Mac in about 2GB of RAM, streaming most of the weights off the SSD 🤓 www.youtube.com/watch?v=vHhe...
Local AI On Apple Silicon uses 7X Less RAM
YouTube video by Better Stack
youtube.com
Wow GitHub MacOS Runner GitHub actions are brutal and costly. On my GitHub Pro account get 3000 minutes/month for Actions but MacOS runner has 300 minutes/month and it blew through my quota in 1 day - doesn't reset for another 30 days! 😮 Signed up for getmac.io 😅
Claude Max $100 + ChatGPT Pro $100 plan token and usage costs via their native OpenTelemetry metrics support exported to Grafana dashboard panels. Claude Code app 24hrs = 219.6 million tokens Codex/ChatGPT app 24hrs = 229.9 million tokens
Google Gmail free accounts getting some Gemini AI on their Google Drive and Sheets and Google Nano Banana Pro on Slides
Upgraded to ChatGPT Pro $100 within ChatGPT desktop app to join my Claude Max $100 plan and other subscriptions 🤓
AI memory demand's impact on Gaming hardware www.youtube.com/watch?v=Eo5w...
The Gaming Market is Starting to Break.
YouTube video by Mrwhosetheboss
youtube.com
Tokens get cheaper but we end up doing more work = overall AI costs increase 🤔
First time seeing Claude Code ask how many dynamic workflow agents to spin up and referenced /config workflow size setting I configured at medium
Ran out of ChatGPT usage limits for the week so can't try ChatGPT Work for first time 😅
OpenAI reduces GPT-5.6 Terra and Luna pricing 80% and 20% respectively after GPT-5.6 improved its own efficiency openai.com/index/advanc... 😎
Advancing the price-performance frontier with GPT-5.6
Explore lower GPT‑5.6 pricing for Luna and Terra—and how OpenAI’s more efficient models help enterprises deploy AI workflows at scale.
openai.com
Nvidia DGX Station with Nvidia GB300 with 748GB unified memory and 400Gbit Network connectivity with 3x PCI-E slots www.youtube.com/watch?v=qV_K...
DGX Station Put a Data Center on My Desk
YouTube video by Alex Ziskind
youtube.com
Nooo ChatGPT Plus out of weekly usage until August 6, 2026 - have 1 reset banked that save for ChatGPT Pro $100 upgrade 🤓
Yes I've posted before but really it is time for ChatGPT Pro $100 upgrade 🤓
ChatGPT Plus subscription GPT-5.6 Sol usage for last 24hrs with weekly resets used = 1.466 billion tokens according to Codex native OpenTelemetry metrics exported to my Grafana dashboard panels 🤓
Claude Fable 5 and GPT-5.6 Sol toasting some tokens over a fire 🔥