iamcoder18

What do you think?

Source: https://collusion.wiki/explorer/page/dse~OAIEquityDec30Raw.h...

I've been using Readest (https://web.readest.com/) but I'll definitely be checking this out.

There's no way GPT 5.5 is better than GPT 5.6 Sol, and gemini-3.6-flash is better than both of those. I wouldn't trust this benchmark at all.

No, I just think it's a great concept and pricing seems good too.

Code.storage is able to handle large scales too: https://code.storage/

+-------------------------+-------------------------------+

| Total storage limit | Unlimited |

| Git request rate limits | Unlimited (Seen: 500+/s/repo) |

| API rate limits | Unlimited (Seen: 15k+/s) |

| Max repos | Unlimited |

| Concurrent writes | 30+ |

| Repository size limit | 32TB |

| Git LFS | Supported |

| GitHub/3rd party sync | Supported |

+-------------------------+-------------------------------+

Check out https://code.storage/

+-------------------------+-------------------------------+

| Total storage limit | Unlimited |

| Git request rate limits | Unlimited (Seen: 500+/s/repo) |

| API rate limits | Unlimited (Seen: 15k+/s) |

| Max repos | Unlimited |

| Concurrent writes | 30+ |

| Repository size limit | 32TB |

| Git LFS | Supported |

| GitHub/3rd party sync | Supported |

+-------------------------+-------------------------------+

Yes, but GitHub enterprise has different uptime (https://us.githubstatus.com/posts/dashboard) and actually does have an SLA.

Edit: it seems GHE is also affected by the outage and isn't much better.

Mine just took screenshots but I had wanted to set it up with accessibility too. That's a cool way to do that.

Just did this locally by giving my AI agent full ADB access to a test phone I have.

It's actually kinda amazing just watching it swipe, open the app launcher, and find issues to fix later.

What models have you been able to run with this?

I've been waiting so long for something amazing to come out of the OpenAI and Cerebras collaboration.

> In our evaluations, GPT-5.6 Sol on Ultrafast mode answered all 2,500 HLE questions in 11 hours and 11 minutes. Claude Fable 5 needed 78 hours and 27 minutes, more than three days of continuous compute, to arrive at the same conclusions. In other words, Ultrafast worked through the frontier of human knowledge in a single working day, achieving comparable accuracy nearly 7× faster.

This is actually insane.

Hopefully the release ultrafast of Terra and Luna too.

> “I think I might need some relaxation time. It feels important to take a breather and find ways to unwind. There’s a lot going on sometimes, and it’s easy to forget to slow down. Maybe I could explore some activities that help clear my mind or consider options like a calming walk, some quiet reading, or just reflecting on things that bring me joy. It’s all about finding that balance, right?”

— GPT-5.6 Terra, reasoning summary, mid-run

This is hilarious

Yeah sorry, I should have clarified that I wish OP's product was open source. mb

This solves my biggest gripe with Mermaid.

It's an amazing tool, but I wish it were open source.

Anything like this for Android or Pixels?

This is so well thought out.

I see. Once again, really cool product! Congrats!

Super cool!

Would you mind giving more insight into how you make the videos (sourcing clips or generating them for Remotion videos, agentic?, and how you handle rendering issues and QA for the Manim videos) and costs?

Absolutely. Haven't heard anyone talk about recently.

This is cool, but has anyone ever needed to rewrite commit authors or dates?

This proves that OpenAI models reason in grug speak to save tokens! I wonder if open models are going to start doing that too to save on reasoning tokens.

I didn't make this, the Orca team did, but I thought it was really cool.

They did one-shot it with Fable.

Repo: https://github.com/stablyai/stockcoaster

Will there be support for lower-end GPUs like RTX 4090s or RTX 6000 Pros in the future?

Cloudflare introduced Code Mode far earlier (https://blog.cloudflare.com/code-mode/)

I've been using Kilo Code (with MiniMax M3) with auto approve (similar to dangerously skip permissions) and I haven't had a single incident.

However, I don't give it long running tasks unsupervised, and I do interrupt it from time to time to give suggestions.

It's hand written and should also work on microcontrollers like ESP32s or Pi Pico 2. Have you tried anything like that yet?

The Pico in particular has enough RAM and supports floating point math.