Pinned Tweet
This V8 is something Astra could only dream of making
Fable 5.2 high
6 hours 43 minutes
$146.54
Zach retweeted
I had 4 models create a game called One-Round. Which model won?
Astra - 37 min, $12.87
Sol-6.1 - 38 min, $1.86
Opus-5.5 - 61 min, $15.60
Fable-5.2 - 1 hr 40 min, $31.21
I had 4 models create a game called One-Round. Which model won?
Astra - 37 min, $12.87
Sol-6.1 - 38 min, $1.86
Opus-5.5 - 61 min, $15.60
Fable-5.2 - 1 hr 40 min, $31.21
Zach retweeted
🚨IF OPENAI HAD FABLE 5.2 🚨
I had fable redesign the codex app. I think it looks way better now.
Zach retweeted
This V8 is something Astra could only dream of making
Fable 5.2 high
6 hours 43 minutes
$146.54
Serious Question why is anyone still paying for an OpenAI subscription? Their models are not near frontier, their usage is worse than Claude, their app is unusable, and their employees troll post on X. Is it pure cope or is there an actual reason you keep paying for a worse product?
Okay here are the things I do not like about OpenAI models in order:
1. They are a pain in the ass to work with and need constant management to ensure they stay on task.
2. They will spin out trying to verify every edge case for something. Makes them good at reviews and reduces bugs at the cost of adding 10,000 lines of test code.
3. They have no taste.
4. The Codex app eats memory and it is bad UI
5. The models are significantly slower than Claude models and this is fact
A lot of people are making statements about using Sol for code review. I agree it’s a great code reviewer and even then it needs to be managed to ensure test code doesn’t get out of hand.
If you can only afford one subscription there is no reason to use Codex. Usage limits are worse, science and research is indistinguishable unless you are at the top 0.001% of a field and Claude models are faster.
If OpenAI releases a better model than Anthropic I will switch my subscriptions. Until then I am Claude all day and you are coping or have a very specific reason to continue using OpenAI models.
Zach retweeted
Serious Question why is anyone still paying for an OpenAI subscription? Their models are not near frontier, their usage is worse than Claude, their app is unusable, and their employees troll post on X. Is it pure cope or is there an actual reason you keep paying for a worse product?
Inspired by @MrTacticalX videos and used his article to make a skill that made this video
teapawt.org/MrTacticalX/status/211…
You all keep asking in my DMs how I make my AI Blender videos and model comparisons...
Paste this into Claude Code and it will set everything up for you.
3
153



