Mia

@MiaAI_lab • 42,659 subscribers

Building with AI & LLMs // Running frontier open models locally

Shorts

Here's one way to get the most from Grok 4.5 ✨ Switching effort to "low" gives strong results on most tasks, near-zero quality loss, big usage savings — plus it's the fastest.

Here's one way to get the most from Grok 4.5 ✨ Switching effort to "low" gives strong results on most tasks, near-zero quality loss, big usage savings — plus it's the fastest.

1,312,268 görüntüleme

People realizing they can run Opus-4.6 level intelligence on their laptop with Qwen3.8-27B. No subscription. No limits. Completely private.

People realizing they can run Opus-4.6 level intelligence on their laptop with Qwen3.8-27B. No subscription. No limits. Completely private.

167,988 görüntüleme

My RTX 5090 is officially doing his part for local AI

My RTX 5090 is officially doing his part for local AI

23,663 görüntüleme

DeepSeek v4.1 Flash on 3x DGX Sparks

DeepSeek v4.1 Flash on 3x DGX Sparks

30,649 görüntüleme

One task, two models. GLM 5.3 Flash running slower at 25 tok/s prose Qwen3.8 Flash running faster at 47 tok/s prose Who will finish the task first?

One task, two models. GLM 5.3 Flash running slower at 25 tok/s prose Qwen3.8 Flash running faster at 47 tok/s prose Who will finish the task first?

27,481 görüntüleme

Until there’s an official Grok Build app, you can still use it from your iPhone or Android with Termius (or any other SSH client). Just install Tailscale on your server, connect to the Tailscale IP in Termius, and run Grok Build from the command line as usual 👇

Until there’s an official Grok Build app, you can still use it from your iPhone or Android with Termius (or any other SSH client). Just install Tailscale on your server, connect to the Tailscale IP in Termius, and run Grok Build from the command line as usual 👇

15,627 görüntüleme

Videos

Anya Rossi

sweetdream.ai

SweetDream.ai•Sponsored•Livecam

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Exclusive private shows

1.2k viewers online

Private Show

Join now for exclusive access

Free preview available • Premium content

DeepSeek V4.1 Flash 1017 on 4x DGX Sparks ⚡️ - Official weights on TensorFold - 8 concurrent streams with thinking on

DeepSeek V4.1 Flash 1017 on 4x DGX Sparks ⚡️ - Official weights on TensorFold - 8 concurrent streams with thinking on

20,248 görüntüleme • 2 gün önce

GLM 5.3 Flash vs DeepSeek V4.1 Flash 1017 Can't believe we're abusing gummy bears to test models, but here we are! Both models were run locally. - GLM 5.3 Flash: my own EXL3 quant (link below) - DeepSeek V4.1 Flash 1017: official quant - Same harness.Both on Max effort. Who won?

GLM 5.3 Flash vs DeepSeek V4.1 Flash 1017 Can't believe we're abusing gummy bears to test models, but here we are! Both models were run locally. - GLM 5.3 Flash: my own EXL3 quant (link below) - DeepSeek V4.1 Flash 1017: official quant - Same harness.Both on Max effort. Who won?

32,512 görüntüleme • 6 gün önce

Whoever made this is a genius 😂

Whoever made this is a genius 😂

246,498 görüntüleme • 2 ay önce

DeepSeek v4.1 Flash via API is almost TOO fast 🤯

DeepSeek v4.1 Flash via API is almost TOO fast 🤯

139,414 görüntüleme • 1 ay önce

This is why DGX Spark prices have sky rocketed

This is why DGX Spark prices have sky rocketed

43,094 görüntüleme • 16 gün önce

That Argon gummy bear video turned out to be a FAKE. I ran the prompt myself, and the results from Gemini 4 Argon is MUCH better. Opus is still better, but Argon is not bad at all. Full prompt below.

That Argon gummy bear video turned out to be a FAKE. I ran the prompt myself, and the results from Gemini 4 Argon is MUCH better. Opus is still better, but Argon is not bad at all. Full prompt below.

27,084 görüntüleme • 11 gün önce

Full GLM 5.3 running on 3x DGX Sparks with TensorFold Currently 24-28 tok/s on prose

Full GLM 5.3 running on 3x DGX Sparks with TensorFold Currently 24-28 tok/s on prose

49,756 görüntüleme • 22 gün önce

Highly recommend to watch the whole thing and pay a special attention to the top-right of the screen.

Highly recommend to watch the whole thing and pay a special attention to the top-right of the screen.

46,133 görüntüleme • 25 gün önce

DeepSeek v4.1 Flash running on 2x DGX Sparks

DeepSeek v4.1 Flash running on 2x DGX Sparks

59,572 görüntüleme • 1 ay önce

Opus 5.5 built me a Fortnite replica in one prompt

Opus 5.5 built me a Fortnite replica in one prompt

44,111 görüntüleme • 1 ay önce

- GLM 5.3 Flash EXL3 - 2x DGX Sparks - TensorFold - 4 concurrent prose streams Pay attention to tok/s and TTFT on each stream. Buttery smooth experience!

- GLM 5.3 Flash EXL3 - 2x DGX Sparks - TensorFold - 4 concurrent prose streams Pay attention to tok/s and TTFT on each stream. Buttery smooth experience!

32,203 görüntüleme • 24 gün önce

Qwen3.8 Flash I really like this model and find it great for frontend. It performs much better than Qwen3.8 27B.

Qwen3.8 Flash I really like this model and find it great for frontend. It performs much better than Qwen3.8 27B.

61,327 görüntüleme • 1 ay önce

I ran this locally with GLM-5.3 Flash EXL3 on my DGX Sparks, and I think it beats both Fable 5.1 and GPT-6 Astra. The crazy part: it was faster to make create this on my sparks than it was with either Fable 5.1 or Astra. Live link here:

I ran this locally with GLM-5.3 Flash EXL3 on my DGX Sparks, and I think it beats both Fable 5.1 and GPT-6 Astra. The crazy part: it was faster to make create this on my sparks than it was with either Fable 5.1 or Astra. Live link here:

48,031 görüntüleme • 1 ay önce

Check out how DeepSeek v4.1 Flash performs on 2x DGX Sparks after the update. Smooth experience!

Check out how DeepSeek v4.1 Flash performs on 2x DGX Sparks after the update. Smooth experience!

39,502 görüntüleme • 1 ay önce

Models now include a proper filtering per device type, and consists 16 different recipes! Includes DGX Sparks, RTX 5090, RTX 6000 Pro, and 16-24 GB GPUs. Let me know what you think!

Models now include a proper filtering per device type, and consists 16 different recipes! Includes DGX Sparks, RTX 5090, RTX 6000 Pro, and 16-24 GB GPUs. Let me know what you think!

30,205 görüntüleme • 1 ay önce

Addicted to Nous Research Hermes Agent's /learn. You can feed it any design and turn it into a reusable skill that makes Hermes design exactly how you want. /learn [URL or image] - extracts design tokens, layout structure, interaction states, component patterns, and a 10-point checklist Next time you want that exact style, just ask Hermes to use the skill it created. Example below 👇

Addicted to Nous Research Hermes Agent's /learn. You can feed it any design and turn it into a reusable skill that makes Hermes design exactly how you want. /learn [URL or image] - extracts design tokens, layout structure, interaction states, component patterns, and a 10-point checklist Next time you want that exact style, just ask Hermes to use the skill it created. Example below 👇

76,206 görüntüleme • 3 ay önce

Testing out MiMo-V2.6 Pro through OpenCode go. It's very slow... mostly around 20-24 tok/s but sometimes even dips below 10 tok/s 🐢

Testing out MiMo-V2.6 Pro through OpenCode go. It's very slow... mostly around 20-24 tok/s but sometimes even dips below 10 tok/s 🐢

22,042 görüntüleme • 1 ay önce

Orinth-1.0-35b vs Qwen-3.6-27b I've asked both to build a single html file of Tetris game. First here's the Orinth 35b I think it did a fine job. Every control works, the game looks fine visually. How did Qwen 3.6 27b did? See post #2

Orinth-1.0-35b vs Qwen-3.6-27b I've asked both to build a single html file of Tetris game. First here's the Orinth 35b I think it did a fine job. Every control works, the game looks fine visually. How did Qwen 3.6 27b did? See post #2

62,696 görüntüleme • 4 ay önce

DeepSeek v4.1 Flash vs GLM 5.3 Flash I asked both models to recreate the Psychonauts 2 and Metroid Prime game menus as closely as possible to the originals. Attached are full-sized videos from both models for both games. All were done in a single shot, both at Max thinking, same harness. Tokens used: DeepSeek v4.1 Flash - Psychonauts 2: 412k - Metroid Prime: 466k GLM 5.3 Flash - Psychonauts 2: 134k - Metroid Prime: 111k That's between 3-4x MORE used by DS4.1F. Who do you think won? IMO GLM 5.3 Flash won by FAR on both.

DeepSeek v4.1 Flash vs GLM 5.3 Flash I asked both models to recreate the Psychonauts 2 and Metroid Prime game menus as closely as possible to the originals. Attached are full-sized videos from both models for both games. All were done in a single shot, both at Max thinking, same harness. Tokens used: DeepSeek v4.1 Flash - Psychonauts 2: 412k - Metroid Prime: 466k GLM 5.3 Flash - Psychonauts 2: 134k - Metroid Prime: 111k That's between 3-4x MORE used by DS4.1F. Who do you think won? IMO GLM 5.3 Flash won by FAR on both.

24,127 görüntüleme • 1 ay önce

I asked Claude Fable 5.1 to create the most beautiful single HTML file (incl. sound). I like it a lot. Full prompt + url below.

I asked Claude Fable 5.1 to create the most beautiful single HTML file (incl. sound). I like it a lot. Full prompt + url below.

30,125 görüntüleme • 1 ay önce