Jonny Q

Building vertical AI products in public.
Sharing lessons from AI SaaS, video, e-commerce, education, Excel automation, and model routing.
What works, what break

[Jonny Q @jonny_quan]

I’m building vertical AI products in public.

Not generic AI news.
Not tool listicles.

I’m testing where AI actually becomes a business:
education, e-commerce, short video, Excel automation, AI SaaS, and model routing.

Sharing what works, what breaks, and what users pay for.

[Jonny Q @jonny_quan]

@hqmank The globe looks great in motion. I’m curious how much of those dozen-plus turns went into the texture and animations.

[Jonny Q @jonny_quan]

In certain scenarios, Gemini offers far more value than either GPT or Claude

[Jonny Q @jonny_quan]

Codex lets me take several days off a week, and here I am waiting for the reset again.

[Jonny Q @jonny_quan]

@hqmank Waiting

[Jonny Q @jonny_quan]

@hqmank Pretty interesting approach to agent memory. The corrections actually carry over to the next run.

[Jonny Q @jonny_quan]

@hqmank Being able to download the weights and choose when to upgrade is a big plus for me.

[Jonny Q @jonny_quan]

@hqmank It's the weekend, and I’m burning through my new quotas ASAP.

[Jonny Q @jonny_quan]

@hqmank Curious how it stacks up against Nano Banana Pro and GPT Image 2

[Jonny Q @jonny_quan]

@hqmank MiniMax has always had a solid track record. Really hyped for what’s coming!

[Jonny Q @jonny_quan]

@hqmank EvoX looks pretty solid. Wondering how it stacks up against Hermes?

[Jonny Q @jonny_quan]

@hqmank DeepSeek’s bill today is straight-up wild. That new peak pricing dynamic really hits hard

[Jonny Q @jonny_quan]

@hqmank With DeepSeek's recent price hike, it just can't compete with the GPT Pro $100 plan anymore. It used to be a fair match at the old price, but GPT Pro also bundled image generation credits, making it way better value

[Jonny Q @jonny_quan]

@hqmank Congrats! Time for me to step up my game and catch up

[Jonny Q @jonny_quan]

@hqmank Hailuo AI is popping off. They're genuinely reshaping the AI video landscape.

[Jonny Q @jonny_quan]

@hqmank MiniMax’s earlier video model was pretty ahead of the curve too, to be fair

[Jonny Q @jonny_quan]

The top-tier models are super accurate right now, but Grok is both fast AND accurate. Such a breeze to work with!

[Jonny Q @jonny_quan]

@hqmank It’s performing great. Fable and GPT-5.6 have been pretty underwhelming lately.

[Jonny Q @jonny_quan]

@hqmank K3’s frontend outputs have leveled up massively—it’s way better than GPT-5.6 now

[Jonny Q @jonny_quan]

Looks like my Claude quotas will reset in the next day or two

[Tibo @thsottiaux]

10M!

New day, new usage reset for paid users of Codex and ChatGPT Work. Lands in the next hour. Enjoy.

[Jonny Q @jonny_quan]

@hqmank Looks like K3 has officially caught up with Fable 3. Absolutely stunning performance.