vibe-coders chat đź”» discord.gg/Z7sSm56rcb
- Tweets 4,225
- Following 893
- Followers 2,848
- Likes 10,347
Pinned Tweet
Tibo, I followed the setup in this post almost exactly. Shortly afterward, Anthropic suspended my account.
I’ve filed an appeal. Full implementation is public here: github.com/alexgetmancom/cla…
@thsottiaux @OpenAIDevs @ClaudeDevs - would appreciate help figuring out what happened and whether this setup is actually prohibited.
If you aren't yet bold enough to install the Codex app, you can stay in the presence of your orange crab and point it at GPT 5.6 Sol. Takes 5 minutes. Kudos to Theo for explaining one of the ways to get this done.
Step 1: Install CLIProxyAPI
Step 2: Connect
Step 3: Define following alias and enjoy claudex
```
alias claudex='CLAUDE_CODE_SUBAGENT_MODEL=gpt-5.6-sol \
CLAUDE_CODE_ALWAYS_ENABLE_EFFORT=1 \
CLAUDE_CODE_MAX_TOOL_USE_CONCURRENCY=3 \
ENABLE_TOOL_SEARCH=false \
claude --model gpt-5.6-sol'
```
If this gets blocked, I owe you a reset.
83
28
1,005
2,031,706
If X had custom reactions, Tibo would see how much people hate resets announced just 3 hours ahead while half the world is asleep
2
10
475
⚡️ Thariq is leaving Anthropic
I think he was the strongest public face of Claude Code, arguably even more than Boris Cherny.
He wasn’t just shipping features and announcing updates. He was genuinely teaching people how agents work.
His deep dives into prompt caching, skills, context engineering, and agent harnesses were some of the best technical writing in AI.
He had a rare ability to explain complex engineering concepts in a way that was actually useful. And he was constantly engaging with the community, sharing insights, and helping people get more out of Claude Code.
This is a huge loss for Anthropic.
I’ve left Anthropic after more than a year of working on Claude Code and seeing it scale larger than I ever imagined.
I have really loved working at Ant and believe deeply in the mission, and most of all, the idea of a race to the top.
The idea that if we work hard and responsibly, we can raise the bar and inspire others to great outcomes. And if we all push each other to be better, things can be radically good. I will carry the spirit of this with me forever.
My colleagues at Anthropic are the most talented, thoughtful, competent, and earnest people I have ever met. I am glad they are at the wheel and trust them deeply.
To the community on X, thank you for everything. You’ve made this the most fun (and sometimes stressful) job I’ve ever had.
For me, it is time for a bit of a break before diving into something new.
If you want more excellent Claude Code content, you should follow @addyosmani , @lydiahallie, @_catwu, @ClaudeDevs and of course @bcherny who started it all.
20
3
159
13,442
Kimi's benchmarks are absolutely mind-blowing. Is there a new sheriff in town?
Introducing Kimi K3.1 Pro: our most capable 3T-class open-weight model.
🔹 1M-token context, 2.8T total / 104B active params (MoE), Multimodal.
🔹 Full-project coding: games in Unreal, Unity & Godot.
🔹 Excels at parametric 3D modeling in Blender & CAD.
🔹 More reliable at finance work: Accurate modeling with hallucination down by 65%.
🔹 Matching the strongest proprietary models across coding, agentic and knowledge benchmarks.
Kimi K3.1 Pro and K3.1 Pro Highspeed (2–3× faster) are now live on Kimi.ai, Kimi Work, Kimi Code, and the Kimi API.
Open weights by Nov 4, 2026.
đź”— API: platform.kimi.ai
đź”— Tech blog: kimi.ai/blog/kimi-k3-1-pro
1
17
972
alex getman retweeted
Introducing Kimi K3.1 Pro: our most capable 3T-class open-weight model.
🔹 1M-token context, 2.8T total / 104B active params (MoE), Multimodal.
🔹 Full-project coding: games in Unreal, Unity & Godot.
🔹 Excels at parametric 3D modeling in Blender & CAD.
🔹 More reliable at finance work: Accurate modeling with hallucination down by 65%.
🔹 Matching the strongest proprietary models across coding, agentic and knowledge benchmarks.
Kimi K3.1 Pro and K3.1 Pro Highspeed (2–3× faster) are now live on Kimi.ai, Kimi Work, Kimi Code, and the Kimi API.
Open weights by Nov 4, 2026.
đź”— API: platform.kimi.ai
đź”— Tech blog: kimi.ai/blog/kimi-k3-1-pro
1,188
2,991
27,685
3,240,907
Kimi K3.1 Pro benchmarks are out!
Of all these benchmarks, I trust Terminal-Bench 4.0 the most.
And the results are impressive:
Kimi K3.1 Pro scores 65.7, just 0.7 points behind the leader, Opus 5.5, while outperforming GPT-6 Astra by 7.5 points.
3
2
21
939
🚨 Kimi K3.1 is out now!
Available outside China via Kimi Code.
Pricing per 1M tokens:
Input / Output: ~$2 / ~$10
I’ll post the benchmarks as soon as they’re available.
5
1
28
1,669
Tibo is resetting Codex limits for the second day in a row. Looks like the compute bottleneck is finally a thing of the past.
Resets are great, of course, and I’ve got plenty of ideas on how to use them. But what I’d really love to see is a new model.
Day 18/ Decisions API is GA, OpenResponses joined the Linux Foundation, we shipped some performance improvements and Uber is going all-in on codex & work. But yes… that’s a reset!
13
47
6,233
alex getman retweeted
Day 18/ Decisions API is GA, OpenResponses joined the Linux Foundation, we shipped some performance improvements and Uber is going all-in on codex & work. But yes… that’s a reset!
1,669
389
10,630
1,868,173
alex getman retweeted
Day 17/ Bet you thought we forgot, but no, it’s reset day today.
Also good time to remind you to get Daybreak Blue access, it only takes 15 mins and you don’t want to have to do it in the middle of an incident.
2,534
560
13,407
2,784,949
Claude Duet, a proactive AI agent from Anthropic, is coming soon.
OpenAI has Dots, xAI has Grok Bot, and there’s also Hermes Agent. Now Anthropic is joining the race.
Looks like we’re moving from the era of coding harnesses to the era of proactive agents that can orchestrate those harnesses on their own.
We’re opening a research preview of Claude Duet, a new app where Claude works as a proactive agent.
It comes to you with what needs your attention, completes tasks you ask of it, keeps working when you step away, and checks with you before it acts on your behalf.
6
1
19
2,059
We’re opening a research preview of Claude Duet, a new app where Claude works as a proactive agent.
It comes to you with what needs your attention, completes tasks you ask of it, keeps working when you step away, and checks with you before it acts on your behalf.
2,329
4,203
54,941
10,528,645
alex getman retweeted
Day 16/
You can now cache for 12h in the API and for logged in users across Codex & Work we keep cache forever and you don’t have any penalty for coming back to an old conversation or resuming after a few hours and are now guaranteed to hit the cache.
You can now choose 12-hour prompt caching in the API.
Persistent agents, like dots, reuse context throughout the day.
You can keep shared instructions, tools, and context cached longer and pay the lower cached-input rate, even when requests are hours apart.
2,613
683
15,476
3,070,345
This is actually a huge update.
Imagine finishing a coding session with 400K tokens of context. You come back a few hours later, the 1-hour cache has expired, and all 400K tokens get processed again without caching, instantly burning through your limits.
Now that’s no longer an issue.
Easily a top 3 update from OpenAI’s sprint for me.
Day 16/
You can now cache for 12h in the API and for logged in users across Codex & Work we keep cache forever and you don’t have any penalty for coming back to an old conversation or resuming after a few hours and are now guaranteed to hit the cache.
7
104
6,188
Tibo reset Codex limit
Now that I have your attention, here’s what I think they’re actually cooking:
1. GPT-6.1 Astra. This one seems pretty obvious by now. The private endpoint has already been leaked, and it looks like we’re getting it this month.
2. Terra is coming back. The Information found evidence that the Tera lineup is being revived. My guess is Tera will become the go-to cheap, efficient model for a while, especially since it feels like everyone’s pretty much forgotten about Luna.
3. That whole coordinates website rumor is complete BS. People keep hyping up some website with coordinates hidden in the code, claiming GPT-6.5 is coming and all that. It’s obviously fake. The guy behind it is literally collecting donations. How are people even falling for this?
13
107
34,622
alex getman retweeted
Day 15/ Nevermind, we're still cooking... and reset. Enjoy.
2,832
461
11,679
1,762,244
OpenAI’s API has revealed three new models:
aurora-green-alpha
aurora-blue-alpha
aurora-red-alpha
About an hour later, three new models appeared on Design Arena: aspen, cypress, and maple.
This strongly suggests OpenAI is testing new image generation models.
Is OpenAI preparing a three-model release soon?
About 2 hours ago, 3 unknown OpenAI models have appeared with the names 'aurora-blue-alpha', 'aurora-red-alpha' and 'aurora-green-alpha'.
About 1 hour after that, 3 new models have appeared on Design Arena with unknown names 'aspen', 'cypress' and 'maple'.
I believe those could be the same exact models tested by OpenAI, but of course I got no proof.
10
6
56
4,590
