Sabitlenmiş Tweet
Theo - t3.gg
62.1K posts
Theo - t3.gg
@theo
Full time CEO @t3dotchat. Part time YouTuber, investor, and developer
San Francisco, CA Katılım Ekim 2016
4.1K Takip Edilen332.7K Takipçiler
Took the day off to play Subnautica 2. Just finished. Very good. Can’t wait for the updates 🫡
English
@avrldotdev Early access but feels very complete and not buggy. Needs a final chapter maybe.
Much better than BZ. Very good sequel to 1. Potential to be better than 1 when it’s finished. Intro pacing was worse but the rest was really good
English
Theo - t3.gg retweetledi
How it started: Claude Code For Real Engineers
How it's going: AI Coding For Real Engineers
English
“Fast and flicker-free” shots fired directly at Anthropic
xAI@xai
An early beta of Grok Build, an agentic CLI for coding, building apps, and automating workflows is now available for SuperGrok Heavy subscribers. Through this early beta, we will improve the model and product based on your feedback. Try it at x.ai/cli
English
@Built_by_APE @iamknownasfesal @mehulmpt Every product I’ve launched was launched with 0 Claude models supported. We always had to delay adding Claude support because of Anthropic’s weird bullshit.
If you’re gonna shit talk you should at least do your research.
English
@iamknownasfesal @theo @mehulmpt Entitled, ungrateful, and disrespectful. His wrappers wouldn't exist without Claude. This crash out is very obviously a weak attempt at strong arming with his "influence." I hope Anthropic holds the line even if it means my broke ass pays more money.
English
The Claude meltdown makes ZERO sense to me honestly. They allow you to use every app in their ecosystem with elevated usage numbers, which is exactly what you want if you are locked into Claude ecosystem (desktop app, design, remote control, etc.)
If you want third party harnesses, there are MANY good models already. Why is it so important to use Opus on some random third party inference? There are models equally good, use them!
English
Theo - t3.gg retweetledi
went through this Claude Sub cancellation thread from Theo
500+ replies, ~70% actual cancellations = 350 people gone (can actually be higher than this)
rough math (assumptions):
- 210 Pro @ $20 = $4,200/mo
- 84 Max $100 = $8,400/mo
- 56 Max $200 = $11,200/mo
$23,800/month. $285K/year. From one tweet.
and this is just the people who replied 💀
Theo - t3.gg@theo
For every person who replies with a screenshot of their cancelled Claude Code plan, I will donate $10 to open source.
English
i have not tried T3Code as there was no provider hook for Kimi before
maybe @theo has added it by now?
English
i would invite people to give this a try as it seems to crash less often than the other things i've tried
won't make anyone's claude code subscription more efficient but otoh it only costs me $40/mo
marketplace.visualstudio.com/items?itemName…
Harrison Kinsley@Sentdex
Apparently an unpopular opinion, but I don't think Anthropic owes anyone heavily subsidized tokens for their third party app.
English
English
What do you mean Theo? In codex, you can have a scripts folder as part of your skill. You can use bash, python, js, or pack something more complex with dependencies.
Maybe you meant something else. Curious if that's the case.
Picture: skill example with scripts (playwright skill by OAI).
Docs:
developers.openai.com/codex/skills
English
Are you still using the CLI versions of your preferred agent instead of desktop apps like Codex App, Conductor, or T3 Code?
Tell me why below. Genuinely curious.
Theo - t3.gg@theo
Just learned it's literally impossible to paste images into Claude Code over SSH. How do you CLI people live like this??
English
Theo - t3.gg retweetledi
I just cancelled my Claude account. I've been using codex, and haven't used Claude in several weeks.
English
@Sentdex @matt_feroz They are not subsidizing the cost of my app? My open source GUI allows you to call Claude Code directly. We implemented this the exact way we were told to.
This is one step away from them charging you more for using Ghostty and requiring you to use the stock terminal instead.
English
@theo @matt_feroz Oh hey nice to meet you too. Sorry I don't think Anthropic should subsidize the cost of your app and that makes you mad.
English
Apparently an unpopular opinion, but I don't think Anthropic owes anyone heavily subsidized tokens for their third party app.
Theo - t3.gg@theo
I can't help but feel personally burned by the Claude Code changes announced today. We put so much work into wrapping the (atrocious) Claude Agent SDK in T3 Code. It was the ONLY path they supported, so we made it work. It was hell. Now our users are getting their rate limits cut by 40x, despite us doing everything right. I listened to the Claude Code team. I had my issues with their direction, but I trusted them and took them at their word. I will never make that mistake again. Until we see significant change, it is safe to assume any statement from an Anthropic employee is a lie on a timer. The rug will be pulled, no matter how many promises are made beforehand.
English
@Sentdex Dishonest framing. There’s a ton of reasonable use cases that Claude Code users relied on that they made impossible. The confusion is insane. They framed a 40x decrease in usage limits like a new feature.
How can you defend this? x.com/mattpocockuk/s…
Matt Pocock@mattpocockuk
I don't know what the fuss is about. Anthropic's rules on using subscriptions are very simple: Claude Code = OK Claude's online platform = OK Agent SDK running in personal software = OK... ish? Agent SDK running in commercial software = NOT OK Claude Code running in CI = ?? Oh, maybe it's not so simple... Agent SDK running in CI = ?? claude -p running in CI = ?? claude -p running in personal software = OK claude -p running on open source software, but run on my personal computer = ?? claude -p running on distributed sandboxes, kicked off by me = ?? Distributing open source software which relies on claude -p, and documenting how to use your subscription with it = ?? A thousand other edge cases = ?? Let me be clear. I have never before experienced, from any developer tool, such a frustrating lack of clarity over the basic terms of usage. I personally asked, 3 weeks ago, and have received nothing but delays. The recent @bcherny announcement did absolutely nothing to clarify things. I say this as someone who just released a Claude Code course - my incentives all align with supporting Anthropic.
English
@Sentdex @matt_feroz Nobody is going to pay 40x more for a better ui you fucking dunce.
English
@matt_feroz Why is thinking Anthropic doesn't owe me or anyone else subsidized tokens shooting myself in the foot? If T3 is so good, why wont you pay the API rate for the tokens? We all know why.
English
@Sentdex @marcospereeira This is a dishonest framing.
If you want a better UI for interacting with Claude code, you were explicitly told to use the agent SDK. They suddenly changed this after we built on top of it.
Now we have to do hacks like this.
x.com/theo/status/20…
Theo - t3.gg@theo
Solution incoming…
English
@marcospereeira It does look to me like it is a benefit. If I'm a claude code user, you're telling me the apps and agents that have been piggybacking on a super cheap api that they get to slam for hella cheap murdering the api perf for me are going away? And I get a $200 monthly api cred? Cool
English
I can't help but feel personally burned by the Claude Code changes announced today.
We put so much work into wrapping the (atrocious) Claude Agent SDK in T3 Code. It was the ONLY path they supported, so we made it work. It was hell.
Now our users are getting their rate limits cut by 40x, despite us doing everything right.
I listened to the Claude Code team. I had my issues with their direction, but I trusted them and took them at their word.
I will never make that mistake again.
Until we see significant change, it is safe to assume any statement from an Anthropic employee is a lie on a timer.
The rug will be pulled, no matter how many promises are made beforehand.
English
1. Claude code is the worst performing harness for Claude. This is established across many benchmarks
2. The UI/UX are awful. Things as simple as pasting images barely work. Their desktop app is a tragedy.
3. My Twitter payouts might seem like a lot of money to you, because you’re broke and don’t employ people. They do not motivate my decision making in any way.
x.com/edwinarbus/sta…
English