Paul Couvert(@itsPaulAi)

Sure. Deeply concerning. Anyway.

Paul Couvert (@itsPaulAi) — Wed, 22 Jul 2026 16:23:07 GMT

Sure. Deeply concerning. Anyway.

Director Michael Kratsios@mkratsios47

We have information that Moonshot AI distilled Anthropic’s Fable for the development of its K3 model. 

To do this they developed a sophisticated internal platform to conduct large scale distillation against U.S. models, allowing them to quickly switch between multiple methods of access to avoid detection. Moonshot AI has also acquired GB300-equipped servers and has accessed GB300s in Thailand, likely to train its AI models.
 
The United States strongly supports the free and fair development of AI, including a thriving competitive ecosystem that spans frontier models, specialized systems, open-source frameworks, and open-weight models. Legitimate AI distillation used to create smaller, more efficient models plays a vital role in this open innovation ecosystem. However, large-scale, covert industrial distillation aimed at stealing proprietary U.S. technology and undermining American research is unacceptable.

🔗 View Quoted Tweet

💬2🔄1❤️12👀5221📊2

⚡ Powered by xgo.ing

Lfg!! Google has just released Gemini 3.6 Flash 🔥 Many are waiting for Pro (me too) but this thing...

Paul Couvert (@itsPaulAi) — Tue, 21 Jul 2026 16:01:04 GMT

Lfg!! Google has just released Gemini 3.6 Flash 🔥

Many are waiting for Pro (me too) but this thing is suprisingly capable

- really fast
- good for frontend tasks
- surprisingly token efficient

And it can even manage to create pretty convincing 3D stuff like this simple Subway Surfers replica!

Pro tip: make sure to enable "extended thinking" in Gemini to improve the results when it comes to coding.

Thanks a lot to Google for the early access 🫶

Your browser does not support the video tag.

🔗 View on Twitter

Logan Kilpatrick@OfficialLoganK

Say hello to Gemini 3.6 Flash, designed to be higher intelligence, more token efficient, and with a new lower price, based directly on developer feedback!

3.6 Flash continues our progress towards models that are deeply usable in real world scenarios!

🔗 View Quoted Tweet

💬6🔄4❤️62👀15291📊12

⚡ Powered by xgo.ing

So Alibaba is about to release an insane open weight model?! Qwen 3.8 max preview is going to be >...

Paul Couvert (@itsPaulAi) — Sun, 19 Jul 2026 13:05:00 GMT

So Alibaba is about to release an insane open weight model?!

Qwen 3.8 max preview is going to be

> second only to Fable 5
> 2.4T parameters
> preview already available

You can already use it today using their (cheap) token plan alongside GLM-5.2, DeepSeek V4 Pro, etc.

Just navigate here int.alibabacloud.com/m/1000415608/

Then select any plan (starting at $4 the first month by claiming the $2 coupon, then $6)

And you can use it in Claude Code, Hermes, anywhere!

💬18🔄8❤️80👀13668📊28

⚡ Powered by xgo.ing

"AI is a public good" → yes "ultimately be provided by the state" → no The biggest risk is seeing...

Paul Couvert (@itsPaulAi) — Sun, 19 Jul 2026 12:19:28 GMT

"AI is a public good" → yes

"ultimately be provided by the state" → no

​The biggest risk is seeing AGI in the hands of only a small group of companies or entities.

​And we avoid that precisely with open-weight models.

Dean W. Ball@deanwball

Some observations on Kimi: 

It's a very good model! I don't think its performance can be explained away by distillation or anything like that. In agentic coding sessions, it seems pretty much on par with the best public models of Q1 2026. In my fairly limited use, it also seemed very token hungry. It's not obvious to me that this model is actually that cheap to run. 

I am personally surprised the Chinese state continues to allow the open sourcing of models this good, given potential risks. To be clear, I *myself* might be fine with models presenting this level of marginal risk being open weight, but I am surprised that China is fine with it. I suspect the reason they are is 75% explained by strategic blindness/lack of AGI-pilledness (the CCP is very Yann Lecun-y in its views of AI). The other 25% or so is their lack of compute for customer inference (making China's open-weight strategy an unintended byproduct of US export controls) and the normal Chinese strategy of aggressive exports. For the companies, as opposed to the government, the decision to open source is partially ideological and partially because they are behind, and they know that very few people would pay for sub-frontier models from China. 

Open-weight models are inherently decelerationist, and I'm continually surprised to see the so-called "accelerationists" so excited about open-weight models. I suspect the reason they are is that they know open-weight models are effectively ungovernable, and they simply like the overall cloak of ungovernability open-weight models create over the whole of AI. It's not a bad strategy; it reminds me of James Scott's recounting of the hill people in "the art of not being governed." Still, in the end, open-weight models deter further AI capex. 

One probable outcome of an open-weight-model-dominant world is full AI communism, which is precisely what China proposes: rather than a market product, AI is a "public good" which will ultimately be provided by the state as a kind of "digital public infrastructure." This future strikes me as a dystopian hellscape, but I've never met an open-weight models advocate who doesn't ultimately concede this is where things end. You'd be surprised how many 'accelerationists' lobbied me, while I was in government, to support an eleven or twelve-figure federally funded data center so that startups could train models at a subsidy and then give them away for free. There was no other way for AI to progress, they said. Perhaps this is the logical end state of things. Nonetheless, I find myself surprised to see supposed accelerationists excited about such an outcome. I think many of them just don't know what they're doing. Many accelerationists do not view the creation and serving of frontier models as a legitimate business. 

I would guess that the Trump Administration will at some point realize that their best strategy here would be to create large amounts of regulatory risk around the use of open-weight Chinese models. You don't need to "ban open source" (one of the dumber motifs of AI policy discussion). You just need to direct every agency to issue soft law that creates FUD. "A Federal Reserve Advisory Bulletin found that there may be backdoors in Chinese AI models." It needn't be that well justified. You just create enough regulatory risk that every regulated enterprise backs off. You probably don't want to create so much regulatory risk that you scare off the hyperscalers from serving Chinese models; this will just drive startups to sketchier providers. There's a happy middle ground here. I'd assume they will do some version of this. 

It's probably true that open-weight models of this capability make the world a bit more dangerous, but not so much more that you'll really notice. At some point the models will be capable enough that you will notice. "A nonliving, invisible, dangerous, and infinitely self-replicating agent escaped from a Chinese lab," you say? Color me shocked.

🔗 View Quoted Tweet

💬12🔄1❤️23👀9559📊10

⚡ Powered by xgo.ing

Highly unexpected Kimi K3 is an open model already competitive with Fable 5 and 5.6 Sol... And eve...

Paul Couvert (@itsPaulAi) — Thu, 16 Jul 2026 19:26:41 GMT

Highly unexpected

Kimi K3 is an open model already competitive with Fable 5 and 5.6 Sol...

And even better on some tasks!!!

- 2.8T parameters
- 1M context window
- Frontier on coding/agentic

Last but not least:

K3 has the same pricing as Sonnet 5 🤯

So you have ~ Fable 5/GPT-5.6 Sol intelligence but MUCH cheaper.

First in frontend design arena as well!

Open source keeps winning

Kimi.ai@Kimi_Moonshot

Meet Kimi K3

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬18🔄16❤️85👀15235📊26

⚡ Powered by xgo.ing

This is bigger than it sounds Thinking Machines has just released the best western open source mode...

Paul Couvert (@itsPaulAi) — Wed, 15 Jul 2026 18:55:05 GMT

This is bigger than it sounds

Thinking Machines has just released the best western open source model 🤯

> open source (apache 2.0)
> multimodal (text/audio/image)
> 41B active parameters

Named 'Inkling' and actually competitive with GPT-5.6 Sol on design arena!

And they are preparing Inkling-Small with 12B active parameters.

Thinking Machines@thinkymachines

Today, we are introducing Inkling.

Inkling reasons efficiently across text, image, and audio modalities. We are making the full weights available.

thinkingmachines.ai/news/introduci…

Available today for fine-tuning on Tinker. Play with it in the Inkling Playground. 🧵

🔗 View Quoted Tweet

💬16🔄13❤️86👀16143📊26

⚡ Powered by xgo.ing

Absolute game changer Anyone can now run a 27B model on a phone 🔥 - Multimodal - Perfect for age...

Paul Couvert (@itsPaulAi) — Tue, 14 Jul 2026 22:39:43 GMT

Absolute game changer 

Anyone can now run a 27B model on a phone 🔥

- Multimodal
- Perfect for agents
- 100% open source

Yesterday it would have required a thousand dollars computer.

Bullish on open source
Bullish on local models

PrismML@PrismML

Today, we’re announcing Bonsai 27B: the first 27B-class model to run on a phone.

Bonsai 27B is the new multimodal flagship of the Bonsai family. Based on Qwen3.6 27B, it brings a new capability tier to local AI: multi-step reasoning, structured tool use, long-context workflows, and coherent agentic loops.

Until now, models in this class have been impractical to deploy locally. A 27B model occupies roughly 54 GB in 16-bit precision, and even a strong 4-bit build is around 18GB - too large for a phone and for most laptops.
Bonsai 27B changes that.

It comes in two variants:
• Ternary Bonsai 27B: 5.9 GB, 1.71 effective bits per weight, optimized for laptop-class quality.
• 1-bit Bonsai 27B: 3.9 GB, 1.125 effective bits per weight, optimized for phone-class footprint.

Everything is open-sourced today under the Apache 2.0 license.

🔗 View Quoted Tweet

💬5🔄8❤️35👀9514📊11

⚡ Powered by xgo.ing

In the last 72 hours we got: - OpenAI GPT-5.6 - SpaceXAI Grok 4.5 - Muse Spark 1.1 - Meta Muse imag...

Paul Couvert (@itsPaulAi) — Fri, 10 Jul 2026 19:56:30 GMT

In the last 72 hours we got:

- OpenAI GPT-5.6
- SpaceXAI Grok 4.5
- Muse Spark 1.1
- Meta Muse image/video
- OpenAI GPT-Live
- Mistral Robostral Navigate

We continue to accelerate.

💬9🔄3❤️27👀7328📊11

⚡ Powered by xgo.ing

I haven't even paid a penny to run the task because it’s currently free on OpenRouter: - Create an ...

Paul Couvert (@itsPaulAi) — Thu, 09 Jul 2026 19:52:42 GMT

I haven't even paid a penny to run the task because it’s currently free on OpenRouter:

- Create an account on OpenRouter
- Generate an API key
- Throw it in Hermes Agent, OpenCode, etc.
- Point it to openrouter.ai/tencent/hy3:fr…

And you’re done! Fantastic work by @TencentHunyuan thanks for the open model!

#Hy3#AIAgents

💬2🔄1❤️8👀2019📊5

⚡ Powered by xgo.ing

This open source model is insanely good Hy3 is the best way to automate any task on your computer w...

Paul Couvert (@itsPaulAi) — Thu, 09 Jul 2026 19:52:08 GMT

This open source model is insanely good

Hy3 is the best way to automate any task on your computer when paired with Hermes agent

> really cheap (even currently free)
> more capable than other open ones
> on par with opus 4.8 on some use cases

And it’s half the size of GLM-5.2 which was the best coding/agentic open model you could use until now.

You can use it for free by following the steps in the post below:

Your browser does not support the video tag.

🔗 View on Twitter

💬4🔄5❤️40👀10397📊16

⚡ Powered by xgo.ing

Here're the links if you want to learn more Hugging Face: https://t.co/Jjx6mpoqo1 Online demo: http...

Paul Couvert (@itsPaulAi) — Thu, 09 Jul 2026 13:44:40 GMT

Here're the links if you want to learn more

Hugging Face: huggingface.co/collections/ro…
Online demo: reactor.inc/lingbot-world-…
Github: github.com/Robbyant/lingb…
Official website: technology.robbyant.com/lingbot-world-…

💬0🔄1❤️2👀1976📊2

⚡ Powered by xgo.ing

Woow things are evolving VERY quickly This open source model can turn any image into a fully intera...

Paul Couvert (@itsPaulAi) — Thu, 09 Jul 2026 13:42:23 GMT

Woow things are evolving VERY quickly

This open source model can turn any image into a fully interactive world 🔥

Worlds are infinite, 720p/60fps... and multiplayer!

Available on Hugging Face and you can try it online (the 1.3B variant deploys on a consumer-grade GPU!)

And you're not limited to just navigation like previous models. You can:

- Shoot
- Cast spells
- Summon new elements

Fully open: both weights and harness!

💬4🔄3❤️15👀6679📊8

⚡ Powered by xgo.ing

Still processing this This open source model is: - Half the size of GLM-5.2 - Competitive w/ GPT-...

Paul Couvert (@itsPaulAi) — Wed, 08 Jul 2026 11:48:20 GMT

Still processing this 

This open source model is:

- Half the size of GLM-5.2
- Competitive w/ GPT-5.5
- Better than DeepSeek V4 Pro

And it costs just a fraction of the current best models.

Tencent cooked.

Open source is evolving so fast and models are MUCH MORE efficient than the closed alternatives.

Tencent Hy@TencentHunyuan

🚀Hy3 is here.

295B MoE. Best in its size class. Rivals trillion-scale flagships.
Reliable and affordable for most agentic usecases. 
Apache 2.0. Friendly for commercial use.
FREE API for 2 weeks →openrouter.ai/tencent/hy3:fr…i

�huggingface.co/tencent/Hy3SL
hy.tencent.com/research/hy3cW4

🔗 View Quoted Tweet

💬31🔄67❤️840👀120118📊161

⚡ Powered by xgo.ing

Europe has the deepest engineering talents on the planet. And I mean it. The only thing missing wa...

Paul Couvert (@itsPaulAi) — Mon, 06 Jul 2026 19:15:12 GMT

Europe has the deepest engineering talents on the planet.

And I mean it.

The only thing missing was ambition and belief.

Glad it just changed.

@viktor__com

Fryd Wiatrowski@frydwia

x.com/i/article/2074…

🔗 View Quoted Tweet

💬4🔄1❤️12👀8510📊4

⚡ Powered by xgo.ing

You should start fine-tuning your own models. For real. You can get better answers with a free ope...

Paul Couvert (@itsPaulAi) — Sat, 04 Jul 2026 19:03:05 GMT

You should start fine-tuning your own models.

For real.

You can get better answers with a free open-source local AI model than with Claude

All you have to do is fine-tune it for YOUR tasks

For many workflows a SLM trained on your specific data can outperform the "general" models from Anthropic or OpenAI.

Among other reasons:

> the model stops giving generic responses
> your sensitive work never touches the cloud
> no usage-based billing
> no expensive API calls
> no usage limits
> works perfectly even offline
> more reliable inside your actual systems

You don’t always need the biggest model.

You need the right one tailor-made for your task.

And sometimes the right model is small, local, private, fast, and trained specifically FOR YOU.

Gemma 4 or Qwen 3.5/3.6 are amazing as a base. Just fine-tune them using unsloth studio.

💬25🔄50❤️227👀18006📊100

⚡ Powered by xgo.ing

Gemini Spark is hands down the most powerful AI tool > no laptop always on > no hard configuration ...

Paul Couvert (@itsPaulAi) — Wed, 01 Jul 2026 21:28:40 GMT

Gemini Spark is hands down the most powerful AI tool

> no laptop always on
> no hard configuration
> no struggle to connect tools

Think of it as openclaw or hermes but on the cloud and easy to use.

It has already automated most of my workflows.

And most importantly:

Gmail, Drive, Docs, Keep, etc. are natively connected... AND you can connect your own MCPs.

Still in beta but already so powerful.

💬16🔄7❤️56👀9584📊26

⚡ Powered by xgo.ing

So AI agents can now buy autonomously the tools they need?! They can pick among 20'000 different on...

Paul Couvert (@itsPaulAi) — Tue, 30 Jun 2026 14:46:38 GMT

So AI agents can now buy autonomously the tools they need?!

They can pick among 20'000 different ones... just based on your prompt.

And the protocol behind it is 100% open (governed by the Linux foundation)

With $1 your agent can buy a run to scrap:

- 350 Google Maps listings
- 430 Instagram profiles
- 110 trending TikTok videos
etc.

Huge!

Your browser does not support the video tag.

🔗 View on Twitter

Apify@apify

Until today, agents could buy about 2,000 tools through x402.

We just 10x'd that to 20,000+ 🚀

In partnership with @coinbase, we’re launching x402 support to give autonomous agents access to the largest marketplace of web automation tools.

No account, API keys, or human in the loop.

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬5🔄6❤️29👀11639📊11

⚡ Powered by xgo.ing

Best local models you can run on a laptop today (And without a $10k Mac Studio) 1. Qwen3.6-27B: by...

Paul Couvert (@itsPaulAi) — Mon, 29 Jun 2026 22:08:38 GMT

Best local models you can run on a laptop today

(And without a $10k Mac Studio)

1. Qwen3.6-27B: by far the best for coding agents
or Qwen3.6-35B-A3B for a faster option

2. Gemma 4 12B: everyday tasks and questions

3. Parakeet 0.6B v3: basically the best voice to text model you can find

4. Gemma 4 E4B: surprisingly good for the size and can run offline on a phone

5. Gemma 4 26B diffusion: the most tok/s you can have with a local model

You should use a quantized version from Unsloth when you're running them on a laptop: the precision/size/speed are fantastic.

LM Studio or the 'raw' llama.cpp are your best options.

Or Google AI Edge Gallery on ios/android.

💬30🔄26❤️281👀24447📊111

⚡ Powered by xgo.ing

That's right btw Utility per watt is the number that will matter A LOT going forward. Basically re...

Paul Couvert (@itsPaulAi) — Thu, 25 Jun 2026 14:55:27 GMT

That's right btw

Utility per watt is the number that will matter A LOT going forward.

Basically real work delivered per dollar of compute. More like a focused AI employee that beats a big general model trying to do everything.

Small and sharp better than a giant generalist.

@viktor__com

Fryd Wiatrowski@frydwia

x.com/i/article/2070…

🔗 View Quoted Tweet

💬7🔄0❤️9👀7516📊5

⚡ Powered by xgo.ing

We're starting soon Already a lot of great people met!

Paul Couvert (@itsPaulAi) — Thu, 25 Jun 2026 07:47:49 GMT

We're starting soon

Already a lot of great people met!

Your browser does not support the video tag.

🔗 View on Twitter

Paul Couvert@itsPaulAi

Landed in Berlin for Google I/O Connect tomorrow

Looking forward to meeting other creators and builders!

Come say hi if you are around 🤝

🔗 View Quoted Tweet

💬1🔄2❤️15👀5377📊2

⚡ Powered by xgo.ing

Landed in Berlin for Google I/O Connect tomorrow Looking forward to meeting other creators and buil...

Paul Couvert (@itsPaulAi) — Wed, 24 Jun 2026 15:48:25 GMT

Landed in Berlin for Google I/O Connect tomorrow

Looking forward to meeting other creators and builders!

Come say hi if you are around 🤝

💬7🔄1❤️30👀13076📊9

⚡ Powered by xgo.ing

You can now see precisely where you're spending your tokens 🔥 This tool is open source and can mon...

Paul Couvert (@itsPaulAi) — Tue, 23 Jun 2026 15:20:47 GMT

You can now see precisely where you're spending your tokens 🔥

This tool is open source and can monitor your agents to:

- Monitor costs
- Catch failures
- Clusters conversations

And compatible with Anthropic, OpenAI, etc.

cesar.wtf@heycesr

Most underrated data source in a company: your AI agent's conversations.

Your agent talks to more customers than any employee, but the data it generates goes nowhere.

@trylatitude changes that, see how:

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬13🔄5❤️24👀7319📊12

⚡ Powered by xgo.ing

There's nothing you can't do with Codex

Paul Couvert (@itsPaulAi) — Fri, 19 Jun 2026 21:57:19 GMT

There's nothing you can't do with Codex

💬12🔄5❤️56👀9366📊17

⚡ Powered by xgo.ing

That's very impressive You can now use an audio-visual AI model in REAL TIME. So you're not just g...

Paul Couvert (@itsPaulAi) — Wed, 17 Jun 2026 15:27:41 GMT

That's very impressive

You can now use an audio-visual AI model in REAL TIME.

So you're not just generating a one-shot video but you can:

> interact with it
> create a real time character
> speak and receive replies

Things are moving so fast

Your browser does not support the video tag.

🔗 View on Twitter

Catnip@catnips_ai

🥇MaineCoon: From Passive Video to Real-Time AI Presence

The first unlimited-duration interactive audio-visual model.

Most AI products today still feel like they live behind a screen.

You type. It answers.
You speak. It replies.

The interaction is still mostly turn-based.

Mainecoon is built around a different idea: AI should not just respond to you. It should feel present with you.

🔗Learn more
Website mainecoon.techs9
Blog mainecoon.tech/blogsT5

🔗 View Quoted Tweet

💬10🔄7❤️33👀45592📊15

⚡ Powered by xgo.ing

Woow Snap has just announced their new augmented reality glasses 🤯 Think of it as a Vision Pro... ...

Paul Couvert (@itsPaulAi) — Tue, 16 Jun 2026 18:03:24 GMT

Woow Snap has just announced their new augmented reality glasses 🤯

Think of it as a Vision Pro...

Except that you're not looking through screens but through real transparent lenses!

You're seeing your REAL environment and can project any virtual element in it.

Use cases are basically unlimited.

Shipping in the US, UK, France for $2195 this fall.

Your browser does not support the video tag.

🔗 View on Twitter

💬6🔄7❤️33👀10312📊12

⚡ Powered by xgo.ing

That’s why open source is vital AI is too important to be left in the hands of closed organizations...

Paul Couvert (@itsPaulAi) — Sat, 13 Jun 2026 13:46:59 GMT

That’s why open source is vital

AI is too important to be left in the hands of closed organizations.

Use these closed models only to create:

- Skills
- Tools
- Datasets
- Workflows

To improve the LOCAL models that YOU own.

Local AI models are now so easy to use and way more capable than you think.

They are now powerful enough to be connected to Codex, Claude Code, etc.

And basically automate or build anything.

Fine-tune them if needed.

💬9🔄5❤️17👀1943📊11

⚡ Powered by xgo.ing

Well here we are Seems that you can now create your ad but also launch the campaigns from a single ...

Paul Couvert (@itsPaulAi) — Tue, 09 Jun 2026 19:54:18 GMT

Well here we are

Seems that you can now create your ad but also launch the campaigns from a single AI tool.

Not just the creative like an image model.

But the whole thing.

Creatify AI@Creatify_AI

Creatify Agent, Wave 2. You watched it make the ad. Now watch it run the whole campaign. It learns your brand. Directs a cinematic spot — no slop. Connects to Meta, TikTok, and Google, and launches on the date you pick. The agent didn't get an update. It got promoted. Drop your product below.

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬5🔄1❤️10👀4011📊5

⚡ Powered by xgo.ing

So you don't need a terminal anymore to use Hyperframes > Connect it to Claude > Ask it for a...

Paul Couvert (@itsPaulAi) — Tue, 09 Jun 2026 16:50:04 GMT

So you don't need a terminal anymore to use Hyperframes

> Connect it to Claude
> Ask it for a video 
> You're done

Zero technical thing.

HeyGen@HeyGen

Hyperframes is now an official @claudeai connector

LLM answers are often dense pages of text that go unread

we partnered with anthropic to fix that, @HyperFrames_ runs as an mcp where you work

docs get skimmed. video gets understood

proof: this video was made with Hyperframes

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬6🔄2❤️7👀3540📊6

⚡ Powered by xgo.ing

You can install it with a single command: winget install Microsoft.IntelligentTerminal The repo is ...

Paul Couvert (@itsPaulAi) — Mon, 08 Jun 2026 21:18:05 GMT

You can install it with a single command:
winget install Microsoft.IntelligentTerminal

The repo is available right here: github.com/microsoft/inte…

It's a fork of Windows Terminal, so you can run it alongside your usual terminal.

💬3🔄1❤️9👀2589📊4

⚡ Powered by xgo.ing

Microsoft has just open-sourced a new intelligent terminal This is a terminal with native agent int...

Paul Couvert (@itsPaulAi) — Mon, 08 Jun 2026 21:16:07 GMT

Microsoft has just open-sourced a new intelligent terminal

This is a terminal with native agent integration in your command line.

It's compatible with any agent CLI like Codex, Claude, Gemini, Copilot, etc.

- Agent status bar
- Agent pane for pair-programming
- Automatic error detection
- Agent management to run multiple agents
- Command palette to kick off an agent task

And this is just the version 0.1 so many things are coming.

💬9🔄5❤️40👀4590📊19

⚡ Powered by xgo.ing

That's massive for local AI Google has just released Gemma 4 QAT and it runs with 3x less memory! ...

Paul Couvert (@itsPaulAi) — Fri, 05 Jun 2026 19:02:46 GMT

That's massive for local AI

Google has just released Gemma 4 QAT and it runs with 3x less memory!

Remember GPT-4o?

Gemma 4 E4B is better and can now run on your phone (!!) with just 2GB RAM.

And Gemma 4 31B (~ Opus 4 level) can now run on your laptop.

💬63🔄132❤️1170👀79549📊293

⚡ Powered by xgo.ing

This NVIDIA model is ridiculously good Same prompt, 10x cheaper. Nemotron 3 Ultra: $0.051 GPT 5.5...

Paul Couvert (@itsPaulAi) — Fri, 05 Jun 2026 13:15:49 GMT

This NVIDIA model is ridiculously good

Same prompt, 10x cheaper.

Nemotron 3 Ultra: $0.051 
GPT 5.5: $0.57

The quality difference between open source and closed models is way smaller than you think.

Your browser does not support the video tag.

🔗 View on Twitter

Paul Couvert@itsPaulAi

Nvidia has just released a REALLY powerful open AI model 🔥

Nemotron 3 Ultra is tailor-made for agentic tasks and coding while being:

- 5x faster than other models
- Up to 30% cheaper
- Available on Hugging Face

Perfect for large codebases, Hermes Agent and more!

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬12🔄33❤️157👀48824📊37

⚡ Powered by xgo.ing

And the benchmarks look promising: It's competitive with models 2x larger... while being cheaper an...

Paul Couvert (@itsPaulAi) — Thu, 04 Jun 2026 15:40:09 GMT

And the benchmarks look promising:

It's competitive with models 2x larger... while being cheaper and faster!

Btw the model isn't only "open weights", but they've also published the data and recipies!

Hugging Face link: huggingface.co/nvidia/NVIDIA-…

💬0🔄3❤️7👀2311📊2

⚡ Powered by xgo.ing

Nvidia has just released a REALLY powerful open AI model 🔥 Nemotron 3 Ultra is tailor-made for age...

Paul Couvert (@itsPaulAi) — Thu, 04 Jun 2026 15:35:49 GMT

Nvidia has just released a REALLY powerful open AI model 🔥

Nemotron 3 Ultra is tailor-made for agentic tasks and coding while being:

- 5x faster than other models
- Up to 30% cheaper
- Available on Hugging Face

Perfect for large codebases, Hermes Agent and more!

Your browser does not support the video tag.

🔗 View on Twitter

💬4🔄7❤️27👀57665📊12

⚡ Powered by xgo.ing

So good to see an alternative this good from a HUGE company like @TencentAI_News You can give it a ...

Paul Couvert (@itsPaulAi) — Wed, 03 Jun 2026 18:50:15 GMT

So good to see an alternative this good from a HUGE company like @TencentAI_News

You can give it a try using their official website: workbuddy.ai

💬0🔄0❤️5👀2755📊1

⚡ Powered by xgo.ing

Most people in the West haven't heard of this AI agent yet. But it’s the no. 1 in China. Tencent Wo...

Paul Couvert (@itsPaulAi) — Wed, 03 Jun 2026 18:49:48 GMT

Most people in the West haven't heard of this AI agent yet. But it’s the no. 1 in China.

Tencent WorkBuddy has millions of users.

- Local folders integrations
- 100s of built-in experts
- Coding and working modes
- Zero terminal, ready in seconds

You can even control it from your phone. From anywhere.

This is probably one of the most powerful app of this type. And now available globally.

Your browser does not support the video tag.

🔗 View on Twitter

💬7🔄9❤️33👀17768📊20

⚡ Powered by xgo.ing

Google has revolutionized AI design with design.md Now we've the equivalent for videos. This frame...

Paul Couvert (@itsPaulAi) — Wed, 03 Jun 2026 16:49:10 GMT

Google has revolutionized AI design with design.md

Now we've the equivalent for videos.

This frame.md tells your agent exactly how to create branded videos.

Finally!

HeyGen@HeyGen

Introducing frame.md, a spec built for videos & motion

design.md kept your brand consistent across screens

but when applied to videos, agents translated it back into webpages and decks

frame.md teaches your agents how to make branded video

turn your design.md into frame.md ↓

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬3🔄1❤️16👀3378📊5

⚡ Powered by xgo.ing

Microsoft has just announced a model better at editing than Nano Banana 2 🔥 - #2 in single-image e...

Paul Couvert (@itsPaulAi) — Tue, 02 Jun 2026 18:36:38 GMT

Microsoft has just announced a model better at editing than Nano Banana 2 🔥

- #2 in single-image edit
- #3 in text to image
- #3 in text rendering

I have been testing it for a few days now and it's surprisingly good!

Only GPT-Image-2 seems a bit better.

Single obvious limitation for now: you can't upload multi-images to edit them for now.

💬7🔄3❤️47👀4562📊11

⚡ Powered by xgo.ing

Open source is going to win We already have an open-weights model competitive with GPT-5.5 and Opus...

Paul Couvert (@itsPaulAi) — Mon, 01 Jun 2026 10:48:06 GMT

Open source is going to win

We already have an open-weights model competitive with GPT-5.5 and Opus 4.7 🤯

MiniMax M3 is genuinely impressive:

- On par w/ GPT-5.5 on SWE Bench Pro
- Neck and neck w/ Opus 4.7 on many
- Better than Gemini 3.1 Pro for coding

But MORE IMPORTANT:

M3 is 10x cheaper to use than both GPT and Opus!!

And the weights will be available on Hugging Face next week. We're so back.

MiniMax (official)@MiniMax_AI

Introducing MiniMax M3: The First Open-Weights Model to Combine Three Frontier Capabilities

- Coding & Agentic Frontier: 59.0% SWE-Bench Pro, 66.0% Terminal Bench 2.1, 34.8% SWE-fficiency, 28.8% KernelBench Hard, 74.2% MCP Atlas
- MiniMax Sparse Attention scales context to 1M
- Natively Multimodal from Step Zero

API: platform.minimax.io
Token Plan: platform.minimax.io/subscribe/toke…
🚀New! MiniMax Code:code.minimax.iol

Weights & Tech Report in ~10 Days

🔗 View Quoted Tweet

💬20🔄5❤️94👀12753📊23

⚡ Powered by xgo.ing

Hear me out Time to update all of them

Paul Couvert (@itsPaulAi) — Wed, 27 May 2026 18:54:49 GMT

Hear me out

Time to update all of them

Tech Actually@techactually

Hear me out too

🔗 View Quoted Tweet

💬8🔄3❤️44👀9086📊10

⚡ Powered by xgo.ing

Well it's accelerating We already had agents with perfect voices and now we can add avatars to them...

Paul Couvert (@itsPaulAi) — Tue, 26 May 2026 17:21:39 GMT

Well it's accelerating

We already had agents with perfect voices and now we can add avatars to them.

This will enable many new use cases...

And much more natural interactions!

LiveAvatar@TryLiveAvatar

Voice agents are getting a face.

LiveAvatar now integrates with LiveKit, Pipecat, Agora, and VisionAgent so the agents you already built can become real-time avatar experiences in a few lines of code.

Keep your stack. Add presence.

Setup guides below ↓

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬5🔄2❤️27👀4653📊7

⚡ Powered by xgo.ing

You can try Qwen3.7-Max with limited time 50% off here: https://t.co/4HOX1DESzD And they also have ...

Paul Couvert (@itsPaulAi) — Sun, 24 May 2026 18:12:50 GMT

You can try Qwen3.7-Max with limited time 50% off here: int.alibabacloud.com/m/1000413223/

And they also have a token plan to use Qwen3.7-Max: int.alibabacloud.com/m/1000413232/

Very impressed by the capabilities but also the speed! Didn’t expect it to be this fast when it comes to browser use.

💬2🔄0❤️13👀4480📊4

⚡ Powered by xgo.ing

So Alibaba has released Qwen-3.7-Max… and it’s really good. You can connect it easily to Hermes Age...

Paul Couvert (@itsPaulAi) — Sun, 24 May 2026 18:12:16 GMT

So Alibaba has released Qwen-3.7-Max… and it’s really good.

You can connect it easily to Hermes Agent or OpenCode to basically replace GPT-5.5 or Opus 4.7.

- 3.3x cheaper output vs Opus 4.7
- 4.0x cheaper output than GPT-5.5

(And  2x cheaper input compared to both)

Everything you need to get started in post below

Your browser does not support the video tag.

🔗 View on Twitter

💬33🔄38❤️356👀40795📊110

⚡ Powered by xgo.ing

Wait so Qwen-3.7-max can beat both GPT-5.5 and Opus 4.7?! While being waaay cheaper: - 9x cheaper ...

Paul Couvert (@itsPaulAi) — Fri, 22 May 2026 10:17:23 GMT

Wait so Qwen-3.7-max can beat both GPT-5.5 and Opus 4.7?!

While being waaay cheaper:

- 9x cheaper than Opus 4.7
- 2x cheaper than GPT-5.5

Very impressive and honestly not expected.

atomic.chat@atomic_chat_hq

Qwen 3.7-max beats Opus 4.7 and GPT-5.5

We tested three frontier models on a real agentic task: write a Tetris bot that plays the game and trains itself. Each model could read its own code, run benchmarks, and rewrite itself across 10 iterations. Then we compared the final bots head to head.

Qwen 3.7-Max: training cost $1.32, bot improvement +56%
Claude Opus 4.7: training cost $12.15, bot improvement +28%
GPT-5.5: training cost $2.85, bot improvement +7%

Qwen won on every dimension - biggest jump, 9× cheaper than Claude, 2× cheaper than GPT. Long agentic loops is where Qwen Max actually delivers.

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬11🔄7❤️90👀15868📊19

⚡ Powered by xgo.ing

Everything announced at the Google I/O 2026 (non-exhaustive): Models: - Gemini 3.5 Flash released -...

Paul Couvert (@itsPaulAi) — Tue, 19 May 2026 23:58:09 GMT

Everything announced at the Google I/O 2026 (non-exhaustive):

Models:
- Gemini 3.5 Flash released
- Gemini 3.5 Pro announced
- Gemini Omni family launched
- Omni Flash multimodal video

Gemini:
- Gemini app full redesign
- Gemini Spark agent launched
- Ask YouTube question answers

Workspace:
- Gemini Live Gmail voice
- Docs Live voice creation
- Keep Live voice creation
- Google Pics announced

AI Studio:
- AI Studio can build Android apps
- AI Studio Workspace integrations
- One-click Cloud Run deployment
- Export projects to Antigravity
- Managed Agents Gemini API

Antigravity:
- Antigravity 2.0 launched
- Antigravity CLI released
- Subagent orchestration enabled
- Antigravity SDK available

Search:
- Search information agents
- Generative UI Search features
- Build apps in Search

Dev tools:
- Android migration agent previewed
- WebMCP web standard proposed
- Chrome DevTools agent support

Other:
- Universal Cart AI shopping
- Gemini for Science launched
- SynthID watermark expanded
- Google Beam lifelike agents

💬13🔄18❤️86👀7484📊29

⚡ Powered by xgo.ing

Gemini Spark might be the most impressive AI announcement of this year. No joke. That's a real auto...

Paul Couvert (@itsPaulAi) — Tue, 19 May 2026 17:52:04 GMT

Gemini Spark might be the most impressive AI announcement of this year. No joke.

That's a real autonomous agent available 24/7 in Gemini

- Can grab context from Gmail, Web, Drive, etc.
- Work autonomously on any task
- Available even from your phone

And we're NOT talking about something coming next year.

But next week!

Your browser does not support the video tag.

🔗 View on Twitter

💬9🔄14❤️72👀8503📊22

⚡ Powered by xgo.ing

And Gemini 3.5 Pro is coming next month!

Paul Couvert (@itsPaulAi) — Tue, 19 May 2026 17:35:25 GMT

And Gemini 3.5 Pro is coming next month!

💬0🔄2❤️10👀2121📊2

⚡ Powered by xgo.ing

Gemini 3.5 Flash has just been announced 🔥 - Better than 3.1 Pro across benchs - 4x faster than G...

Paul Couvert (@itsPaulAi) — Tue, 19 May 2026 17:34:47 GMT

Gemini 3.5 Flash has just been announced 🔥

- Better than 3.1 Pro across benchs

- 4x faster than GPT-5.5 or Opus 4.7 (!!)

- Also WAAAAY cheaper than them

💬3🔄3❤️26👀3984📊6

⚡ Powered by xgo.ing

We're so ready! Main conference of the Google I/O in 25 mins now 🔥

Paul Couvert (@itsPaulAi) — Tue, 19 May 2026 16:36:32 GMT

We're so ready!

Main conference of the Google I/O in 25 mins now 🔥

💬11🔄0❤️58👀5809📊14

⚡ Powered by xgo.ing

There are 5x more Windows users than Mac users But you can officially control Codex from your phone...

Paul Couvert (@itsPaulAi) — Mon, 18 May 2026 23:32:21 GMT

There are 5x more Windows users than Mac users

But you can officially control Codex from your phone only if you're using macOS.

Turns out the feature exists but is hidden on the Windows version.

So here's how to activate it to control Codex from your phone:

1. Open Codex Windows and go to the settings
2. Under 'Configuration', click on 'Open config.toml'
3. In the file, add these 2 lines and save it:

[features]
remote_control = true

4. IMPORTANT: now you need to make the config.toml read-only.

- Open its location in Windows (usually C:\Users\[your username]\.codex) 

- Right click on it to open the properties

- Check the 'Read-only' box, click apply, and restart Codex.

Now your phone will be able to connect to your Codex Windows!

💬62🔄41❤️288👀45669📊139

Paul Couvert(@itsPaulAi)

Sure. Deeply concerning. Anyway.

Lfg!! Google has just released Gemini 3.6 Flash 🔥 Many are waiting for Pro (me too) but this thing...

So Alibaba is about to release an insane open weight model?! Qwen 3.8 max preview is going to be >...

"AI is a public good" → yes "ultimately be provided by the state" → no ​The biggest risk is seeing...

Highly unexpected Kimi K3 is an open model already competitive with Fable 5 and 5.6 Sol... And eve...

This is bigger than it sounds Thinking Machines has just released the best western open source mode...

Absolute game changer Anyone can now run a 27B model on a phone 🔥 - Multimodal - Perfect for age...

In the last 72 hours we got: - OpenAI GPT-5.6 - SpaceXAI Grok 4.5 - Muse Spark 1.1 - Meta Muse imag...

I haven't even paid a penny to run the task because it’s currently free on OpenRouter: - Create an ...

This open source model is insanely good Hy3 is the best way to automate any task on your computer w...

Here're the links if you want to learn more Hugging Face: https://t.co/Jjx6mpoqo1 Online demo: http...

Woow things are evolving VERY quickly This open source model can turn any image into a fully intera...

Still processing this This open source model is: - Half the size of GLM-5.2 - Competitive w/ GPT-...

Europe has the deepest engineering talents on the planet. And I mean it. The only thing missing wa...

You should start fine-tuning your own models. For real. You can get better answers with a free ope...

Gemini Spark is hands down the most powerful AI tool > no laptop always on > no hard configuration ...

So AI agents can now buy autonomously the tools they need?! They can pick among 20'000 different on...

Best local models you can run on a laptop today (And without a $10k Mac Studio) 1. Qwen3.6-27B: by...

That's right btw Utility per watt is the number that will matter A LOT going forward. Basically re...

We're starting soon Already a lot of great people met!

Landed in Berlin for Google I/O Connect tomorrow Looking forward to meeting other creators and buil...

You can now see precisely where you're spending your tokens 🔥 This tool is open source and can mon...

There's nothing you can't do with Codex

That's very impressive You can now use an audio-visual AI model in REAL TIME. So you're not just g...

Woow Snap has just announced their new augmented reality glasses 🤯 Think of it as a Vision Pro... ...

That’s why open source is vital AI is too important to be left in the hands of closed organizations...

Well here we are Seems that you can now create your ad but also launch the campaigns from a single ...

So you don't need a terminal anymore to use Hyperframes > Connect it to Claude > Ask it for a...

You can install it with a single command: winget install Microsoft.IntelligentTerminal The repo is ...

Microsoft has just open-sourced a new intelligent terminal This is a terminal with native agent int...

That's massive for local AI Google has just released Gemma 4 QAT and it runs with 3x less memory! ...

This NVIDIA model is ridiculously good Same prompt, 10x cheaper. Nemotron 3 Ultra: $0.051 GPT 5.5...

And the benchmarks look promising: It's competitive with models 2x larger... while being cheaper an...

Nvidia has just released a REALLY powerful open AI model 🔥 Nemotron 3 Ultra is tailor-made for age...

So good to see an alternative this good from a HUGE company like @TencentAI_News You can give it a ...

Most people in the West haven't heard of this AI agent yet. But it’s the no. 1 in China. Tencent Wo...

Google has revolutionized AI design with design.md Now we've the equivalent for videos. This frame...

Microsoft has just announced a model better at editing than Nano Banana 2 🔥 - #2 in single-image e...

Open source is going to win We already have an open-weights model competitive with GPT-5.5 and Opus...

Hear me out Time to update all of them

Well it's accelerating We already had agents with perfect voices and now we can add avatars to them...

You can try Qwen3.7-Max with limited time 50% off here: https://t.co/4HOX1DESzD And they also have ...

So Alibaba has released Qwen-3.7-Max… and it’s really good. You can connect it easily to Hermes Age...

Wait so Qwen-3.7-max can beat both GPT-5.5 and Opus 4.7?! While being waaay cheaper: - 9x cheaper ...

Everything announced at the Google I/O 2026 (non-exhaustive): Models: - Gemini 3.5 Flash released -...

Gemini Spark might be the most impressive AI announcement of this year. No joke. That's a real auto...

And Gemini 3.5 Pro is coming next month!

Gemini 3.5 Flash has just been announced 🔥 - Better than 3.1 Pro across benchs - 4x faster than G...

We're so ready! Main conference of the Google I/O in 25 mins now 🔥

There are 5x more Windows users than Mac users But you can officially control Codex from your phone...

"AI is a public good" → yes "ultimately be provided by the state" → no The biggest risk is seeing...