Quantinine

[Jéferfon Menezes @JefinhoMenes]

COMO ASSIM BICHO

O modelo de IA da PREFEITURA DO RIO bateu os modelos da Qwen e da Deepseek nos benchmarks e virou um dos melhores modelos de IA de código aberto ????????

[Quantinine @quantinine]

Wiser words have not been said

[𝗭𝗲𝗻 𝗠𝗮𝗴𝗻𝗲𝘁𝘀 @ZenMagnets]

@VictorTaelin It means it's time to buy a gpu

[GPTware @GPTWare]

Fable is likely just Qwen4.x - 3-4 months in advance. If you want Fable locally for free in 4 months, buy a GPU.

[Cohere @cohere]

When you rent your artificial intelligence, you have no control, and no choice. This is why sovereignty and ownership matters.

Whether it means using your own hardware, open source, or deep customization. Own your AI, own your future.

[Quantinine @quantinine]

Gregs open for 10 minutes every day

[GSM 🍑💹 @gsm739]

The UK should be expending every available iota of energy on persuading Anthropic to relocate to the UK. Give them their own nuclear power plant. I am happy to only have electricity in the evenings.

[Quantinine @quantinine]

@elliotarledge it will not drop. by end 2026 you can expect to pay double that of what you had to pay in march 2026

[Andreas Kirsch 🇺🇦 @BlackHC]

The UK & Europe need a frontier AI lab and we need a lot more compute in Europe as a whole

[Chris Poensgen @Poensgi]

@aidangomez we need a marketplace of frontier labs from different schools of thought

[Quantinine @quantinine]

Imagine

[Steve💙🇨🇦 @xyster]

It feels like it's only a matter of time before we won't be allowed to buy our own AI hardware.

Cherish what you have

[Quantinine @quantinine]

@jxnlco @steipete I only see fashion posts

[Quantinine @quantinine]

“Can you guys wait until we catch up?”

[Kanishka Narayan MP @KanishkaNarayan]

Access to @AnthropicAI’s latest models has been paused for all customers, including in the US and UK.

The main lesson: as we debate the future of national security and technological sovereignty, access to AI capabilities is crucial.

That's why this Govt is the first to set up dedicated funding for our AI industry through @UKSovereignAI unit. And just this week, we announced £1.1bn for our AI chip industry.

[MiniMax (official) @MiniMax_AI]

M3 would never 🙂‍↔️

As a matter of fact, the weights are now open, too.

https://t.co/g4Ybfa2kWH

[Quantinine @quantinine]

@calcarinus @NoetekCo @Kimi_Moonshot @basedjensen @TheAhmadOsman Start with qwen nvfp4 instead

[Quantinine @quantinine]

Smaller than expected.

Very promising!

[MiniMax (official) @MiniMax_AI]

MiniMax M3, Open-Weight, Now On Hugging Face , with only ~428B parameters and ~23B activated parameters

Weights:
https://t.co/g4Ybfa2kWH
MiniMax Sparse Attention:
https://t.co/HcTlWRotG3

[Quantinine @quantinine]

@NoetekCo @calcarinus @Kimi_Moonshot @basedjensen @TheAhmadOsman Don’t forget cache and multi request for anything agentic. ++

[Quantinine @quantinine]

@calcarinus @Kimi_Moonshot @basedjensen @TheAhmadOsman You need a minimum of 8x RTX PROs Blackwell

[Espen JD @Snixtp]

DiffusionGemma 26B A4B NVFP4 on RTX Pro 6000

Decode: 852 tok/s

Prefill peak: >50,000 tok/s

Diffusion models is the future, I want more

[Quantinine @quantinine]

@kreoxi @OpenAI @OpenAIDevs It was really fun! The competition was top tier, lots of GPU churning.(You should join if they do announce a new one)

[Quantinine @quantinine]

We need another one! @OpenAI @OpenAIDevs

[OpenAI @OpenAI]

Are you up for a challenge?

https://t.co/GNryIDhnut

[Quantinine @quantinine]

King maker moves

[Google Gemma @googlegemma]

Meet DiffusionGemma!

An experimental open model that explores a fast approach to text generation, released under an Apache 2.0 license.

Moving beyond sequential, token-by-token processes to generate entire blocks of text simultaneously. Here’s what’s new with DiffusionGemma: 👇