taner senyurt

[Gabriel Olson @Gabriel_Olson]

DuckTales premiered 39 years ago today.
I loved that show. Not for the gold. For the idea that adventure meant going somewhere wild with people you love.

It was Indiana Jones for kids.

Nobody's making the movie, so I made the trailer. Sound on. 🦆

[Diogo Almeida @CompleteSkeptic]

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?

I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev

• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions

AFAICT the shortest path to AI-based economic revolution

[taner senyurt @turshu]

@MonkJonSmith @BobGolen

[taner senyurt @turshu]

@antirez @ivanfioravanti i was use for a while from the feature branch it is pretty amazing thank you guys appreciated

[0xSero @0xSero]

New deepseek will release tomorrow!

Flash beats current pro at everything

Deepseek still in the game

[David Dalcu @ddalcu]

MLX-Serve 26.9.2 is out. Fixes, Polish, Speed.

If you have 96-128GB Ram, 👑 Qwen Flash Next is the king of local models for now. Without KV8 Cache + M5Max, you might see it go above 150 Tok/s.

https://t.co/gKpzkuh03M

[taner senyurt @turshu]

@ddalcu @Beamsters1 Stunning result, I will test it as soon as I have time. in this case low power mode pretty usable without listenning fan noise. Thanks for your hard work

[Jun Song @jun_song]

Korea is giving free, unlimited AI to its entire population of 50M.

The plan seems to be serving models built by domestic Sovereign AI startups straight from local data centers.

These models will likely trail behind free GPT at first. But the real game is securing usage data to scale domestic AI capabilities.

More countries should take notes and try this.

[taner senyurt @turshu]

@ivanfioravanti @kernelpool @ddalcu @Spangler3000 if its ready i can test it just give me the branch name

[taner senyurt @turshu]

@ozansihay The Crow

[taner senyurt @turshu]

@0x0SojalSec there is no such repo exist currently gives 404

[Daniel @imnotchalk]

Qwen3.8 27B is now available on @cerebras Shared Tier at ~1500 tok/s. 🚀

It's so fast that the time I spent on writing the prompt was longer than the time it spent to summarize my codebase.

[Oğuzhan Atsız @oguzzatsiz]

Sorunu olan herkesin sesi oldu ama şimdi sesi duyulmuyor. Oğuzhan Uğur neden tutuklu?

[taner senyurt @turshu]

@jetbrains and also i think default model needs vision support too

[taner senyurt @turshu]

@jetbrains is it possible to change local model to qwen 3.8 flash version or is there any plan to change default model to it

[taner senyurt @turshu]

@AlicanKiraz0 ileride yapılacak iyilestirmelerle iyi calismasina calisir ama fan sesi ve sicaklik artisi ile makineyi ayni anda kullanmak pek makul degil low power modda 40-50 token verirse super olur, llm icin external makine alip uzaktan kendi makinende kullanmak en ideali gibi gozukuyor

[Qwen @Alibaba_Qwen]

⚡Meet Qwen3.8-Flash, a multimodal MoE and an early preview of the Qwen4 architecture, now open-weight!

The production version Qwen3.8-Flash will be available soon via QwenCloud API at just $ 0.16/1M input tokens and $ 0.47/1M output tokens.

125B parameters + 51B N-gram embeddings, with just 6B activated per token. Unmatched cost-efficiency.

What's new: 🥳
- Next architecture: GDN + QSA hybrid attention, Gated Residual, N-gram Embedding & Muon optimizer, serving as a precursor to the architecture used in Qwen4.
- Dramatically lower training and inference costs: trained at just 1/9 the cost of Qwen3.7-Plus, while outperforming it across the board with especially strong gains in coding and office tasks.
- Strong performance: scoring 58.7 on DeepSWE 1.1, 62.5 on SWE-bench Pro, 73.9 on CoWorkBench, 84.5 on AndroidWorld, and 95.7 on MathVision (with CI).
- 262K native context, extensible to 1M with YaRN.

We’re also releasing the weights for Qwen3.8-Flash-Next, giving the community an early look at the new architecture we’re exploring for Qwen4.🚀

We can't wait to see what you build with Qwen3.8-Flash!👀👇
- Blog: https://t.co/M5hYypFLgJ
- Technical Report: https://t.co/IF0gObIkQO
- Hugging Face: https://t.co/6ow8QVAABt
- ModelScope: https://t.co/tDOn2jNuFG

[Nick @NickDevFE]

img2threejs 1.5.1 is out 🚀

Checkout: https://t.co/w7cBzRmENO

This release introduces a new approach to image → 3D → Three.js.

Image → Tripo → img2threejs → reconstructed geometry → Three.js

Instead of using the generated 3D model as the final asset, img2threejs uses it as a reference to rebuild the character into a representation that can be generated and rendered directly in Three.js.

No Blender workflow, and no GLB required as the final output.

The animation in the demo is still a bit rough because this release is focused on reconstruction, not animation. I’ll be improving that in the next release.

Still experimental, but this is a big step toward where I want img2threejs to go.

[taner senyurt @turshu]

@MiaAI_lab its great for game creation too, it creates perfect games in nearly one shot