@VLEAnderson_ @DutchLemonade @theo There is an option in the API Keys settings that lets you fallback to BYOK when you exceed our context limits.
English
yash
144 posts
yash
@ysinghdotus
HS dev who also plays basketball | intern @t3dotchat
Katılım Mart 2023
1K Takip Edilen633 Takipçiler
Cross model and cross session comparison of occurrences of strings generated by LLMs, well within rights considering the context in which I signed up for your service, review our previous x.com interactions.
Theo - t3.gg@theo
@VLEAnderson_ Why are you sending a dozen PDFs to the most expensive model that we offer? What exactly is the goal here? We have a PR merging soon that will explicitly reject requests like this 🙃
English
English
@ysinghdotus @theo @OfficialLoganK @r_marked @jullerino Can DM you the thread ID, but your DMs are not open.
English
T3 Chat is $8 and solves all of this. Makes Nano Banana actually usable
Klaas@forgebitz
i am going to sound like a boomer but why is it so hard to run nano banana? like i google for it, 20 other sites, finally find gemini turn on "nano banana" but it's not the latest one get some garbage images have to figure out how to run the "pro" version? turns out it can't edit people anymore 90% of use cases for image enhancement/edits gone..
English
@theo @OfficialLoganK @aryankumar877 @r_marked @jullerino ive def seen gemini 3 hang sometimes, but given that this specific msg has been stuck for 24hr+, im guessing that the request got lost and left convex in a waiting state. happy to confirm if this is the case @aryankumar877 (you can dm me the thread id)
English
@OfficialLoganK @aryankumar877 @r_marked @jullerino Iirc it might be a bug around parallel tool calling (tagging in @ysinghdotus for more context here)
English
@reach_vb You can also use mermaid.live, it has syntax highlighting and all
English
I love LLMs because I can just prompt them to build simple disposable tools that I'd otherwise not - simple mermaid visualiser
it even hooked it up to github pages - all with one prompt; ty codex!
English
Trying to find an answer on XDA, but there's so much things :) @ysinghdotus do you still have your xda dump? is there a chance of opensourcing that knowledge?
English
what's the best android phone for rooting/hacking around etc?
like the best choice from the modding community/xda
is it pixel 6?
(i'm absolute android noob, always had iphones, but i want to have a hacking device that i can play with, reverse engineer apps, sniff BT, etc)
English
over-analyzed a bit here. sonoma-dark makes a novel mistake on the hardflip test (calls it a "frontside 180 hardflip" which would be a different trick)
Adding to negative answer check, but did a quick search first. I've never seen any model with this incorrect answer and Dusk hit it 15 separate times.
Really really interesting.
English
Two new models dropped. I have legit no idea who made these models. I don't even know what the difference is between them. Anybody have hints for me?
English
@spacemonster @theo @jullerino We don’t auto highlight if you paste in code, if you want syntax highlighting you’ll have to wrap it in backticks and label the language used
English
English
Small T3 Chat change that I'm really hyped about - the "edit" button now transforms your message into a full input box with all the expected controls :)
English
@scholzmx @theo I have been using this MLX implementation which has been working well for me: github.com/senstella/para…
English
@ysinghdotus @theo It also requires an Nvidia GPU, so somewhat limited.
English
@ysinghdotus @theo Never heard about them, exactly how they are better than whisper?
English
English
i need someone smarter than me to explain how to store customer API keys securely
building "bring your own keys" functionality seemed simple and now i am researching AWS KMS at 3 am
English
What is the difference of mlx versus llama.cpp/gguf? Never seemed to make it much better than the other but gguf runs everywhere
Stavros Kassinos@KassinosS
If the rumors are true and the MLX team is considering leaving, it would mark one of Apple’s worst strategic missteps at a critical moment. Killing MLX wouldn’t just derail Apple’s AI trajectory—it would undercut the entire Studio line and the M-series chips. @tim_cook, nurture the MLX team and let Studio bloom. If Apple drops MLX now, it’s not just a fumble—it’s a faceplant. This isn’t a side project—it’s a core enabler of cutting-edge, on-device AI. @tim_cook, this is your moment to double down. Backing MLX powers the Studio, strengthens Apple’s AI position, and inspires a developer community that’s already all in.
English
good thing i have a fast connection
Theo - t3.gg@theo
ANNOUNCING THE FIRST EVER CLONEATHON Build an open source clone of T3 Chat, win up to $5,000 Deadline is next Wednesday. Good luck and have fun nerds 🫡
English
@varo9v @t3dotchat yeah, i sadly haven't gotten the chance to update it :(
English
@r_marked @t3dotchat They all have rest APIs and I’m willing to stand up shitty rest to MCP servers to get the job done and iterate on them over time
English
Please help me bully Theo into adding this into @t3dotchat
Dev Agrawal@devagrawal09
@theo Remote MCP
English
I think the people following this account may be braindead?
stupid tech takes@stupidtechtakes
submission by @luciascarlet
English
@getorga @tingamer250 @drdroo @TimSweeneyEpic They could leave it unsigned, but the recent updates have made it a little too confusing for normies to figure out how to run unsigned apps. But tbf the same was the case for the EU iOS versions.
English
We will return Fortnite to the US iOS App Store next week.
Epic puts forth a peace proposal: If Apple extends the court's friction-free, Apple-tax-free framework worldwide, we'll return Fortnite to the App Store worldwide and drop current and future litigation on the topic.
Tim Sweeney@TimSweeneyEpic
NO FEES on web transactions. Game over for the Apple Tax. Apple’s 15-30% junk fees are now just as dead here in the United States of America as they are in Europe under the Digital Markets Act. Unlawful here, unlawful there. 4 years 4 months 17 days.
English