Uday Yatnalli (@udaysy) header
Uday Yatnalli (@udaysy) profile picture

building real businesses with humans & AI agents 👨‍🎓 https://teapawt.org/t.co/vyDEeyKgDV $20k/mo 🇩🇪 https://teapawt.org/t.co/09IrPaxL2x $1.2k/mo 💻 https://teapawt.org/t.co/slhBDQlaKK $10k/mo 👷‍♂️ https://teapawt.org/t.co/ejOXm4q95j

Joined September 2015
Photo by Uday Yatnalli (@udaysy)Photo by Uday Yatnalli (@udaysy)Photo by Uday Yatnalli (@udaysy)Photo by Uday Yatnalli (@udaysy)Photo by Uday Yatnalli (@udaysy)Photo by Uday Yatnalli (@udaysy)Photo by Uday Yatnalli (@udaysy)Photo by Uday Yatnalli (@udaysy)Photo by Uday Yatnalli (@udaysy)Photo by Uday Yatnalli (@udaysy)
my twitter account is 10 years old had 67 followers until last week in that time i: - finished masters at TU Munich - quit corporate after 2 years - bootstrapped 3 companies from India - built a team of 25 - shipped products used by 100K+ students posted maybe 50 tweets total biggest mistake? not building in public from day 1 fixing that now. following everyone whos on the same journey
16
1
51
7,611
i pay openai $200 a month for chatgpt pro. i redeemed my codex usage reset around 7am today and my weekly limit ran out at 8:46 43% of that week went to 5 codex answers. everything below comes straight from codex's own logs, so @OpenAI @OpenAIDevs can check it on their side
4
1
416
openai's own usage api afterwards shows limit_reached true with 0 resets left. the window opened at 01:25 utc when i redeemed the reset, and the limit was hit at 03:16 utc
1
51
for anyone at openai who wants to check: session 01a16168-f448-7c91-9d5e-77a87e9098ae, codex cli 0.163.0, gpt-6-astra, 22 oct 02:31 to 03:16 utc. all 5 response ids are in the image
28
called it..
pretty sure this is zhipu's next GLM. same tokenizer, same z.ai error messages, and it overthinks exactly like 5.3 does good for quick stuff, gets lost on long agent runs.
202
pretty sure this is zhipu's next GLM. same tokenizer, same z.ai error messages, and it overthinks exactly like 5.3 does good for quick stuff, gets lost on long agent runs.
Ox Alpha (stealth model) is free for the next week - 1M Context - Multi-modal - Zero Data Retention Generous rate limits, near unlimited usage We have capacity for 100T tokens per day, lets see what you can do
1
1
1
468
wow @WisprFlow now has recorder & notetaker! much needed and greatly enhances my workflow with ai agents. kudos @tankots!
1
1
2
554
you know your customers!
88
love this promo..even though my experience is quite the opposite working with ai agents, spending more time on my mac more than ever.
Meet Qwen3.8-Max: A New Bar for Coding and Cowork.
1
213
same, but never felt better than fable, faster maybe.
Opus 5 experience lately Has anyone noticed this keeps happenin to Claude models?Opus 4.8, Fable 5, Sonnet 5 now Opus 5 Over time, the reasoning gets worse, hallucinations increase & the model gets nerfed Why it's always Claude but not nearly as often with GPT Grok or Gemini
172
it’s baffling to me that such premium priced software still doesn’t have official cli @airtable
We've partnered with @OpenAI to bring Airtable directly into @ChatGPT. Build fully customized workflows and manage your data without ever leaving your chat. → Connect your bases in seconds → Generate new tables and structures mid-conversation → Turn unstructured chat ideas into live workflows instantly Connect:chatgpt.com/plugins/plugin_a…
167
if claude did 80% price drop for haiku, would you be so hyped?
Absolutely freaking insane. OpenAI is dramatically escalating the AI price war. Faster inference usually carries a disproportionately higher price: providers sacrifice batching efficiency and reserve scarce compute to reduce latency. But GPT-5.6 Sol Fast delivers up to 2.5x the speed for only 2x the price, giving customers a 25% better speed-to-price ratio with exactly the same intelligence!! At the same time, OpenAI cut GPT-5.6 Luna’s price by 80% to just $0.20/$1.20 per million tokens, while Terra dropped 20% to $2/$12. Luna is now 25x cheaper than Sol. OpenAI is attacking both ends of the market simultaneously: near-commodity intelligence at the bottom and frontier intelligence with far better latency economics at the top. The AI price war just entered its next phase. And OpenAI just declared war at china’s cheap intelligence.
1
159
great, now given all the "suffering" we had using luna at 100% cost, can we have our reset now?
We kept ourselves busy with GPT-5.6 Sol. It's pretty good at things. Rolling out - 80% lower prices on Luna - 20% lower prices on Terra - faster speeds on GPT-5.6 Sol in /fast mode - ~ 10X cheaper auto-approve mode ("review for me" in the app), which now uses Luna to prevent many high risk actions from the main agent Towards intelligence too cheap to meter and many more breakthroughs ahead.
168
so all this hype was for those 43 people using gpt luna?
This week is all about intelligence too cheap to meter. Tomorrow we ship again.
8
650
Uday Yatnalli retweeted
Replying to @jerryjliu0
the full loop: google it, dont read it, ask gpt, dont read that either, ask claude to explain what gpt said
2
3
448
Uday Yatnalli retweeted
Replying to @naval
un-investable for VCs looking for 100x, sure. still the best business model for a bootstrapped founder tho. the difference is whether your moat is the code itself or the domain expertise and customer relationship underneath it. ai eats code, it doesnt eat trust
11
7
1
287
37,495
every developer in 2026
1
6
478
"use one AI tool for everything" is the new "use one programming language for everything" pick the right model for the job
2
10
354