Loading video...

Video Failed to Load

Go Home

mlx_lm.server + Qwen3 Coder Next 6bit + OpenCode + M3 Ultra = a pretty capable and very fast local coding setup. Not sped up:

awnihannun's profile picture

Awni Hannun

44,822 subscribers

66,619 views • 4 months ago •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

MLX MiniMax 2.5 running LOCALLY on a single M3 Ultra 512GB! Writing a poem on LLMs at 6bit quantization! 🔥 Let's start some coding, context and distributed tests! Generation: 40.2 tokens-per-sec Peak memory: 186 GB
0:11

MLX MiniMax 2.5 running LOCALLY on a single M3 Ultra 512GB! Writing a poem on LLMs at 6bit quantization! 🔥 Let's start some coding, context and distributed tests! Generation: 40.2 tokens-per-sec Peak memory: 186 GB

Ivan Fioravanti ᯅ

226,052 views • 3 months ago

someone submitted a pr that added "beast prompt" to opencode for gpt models it makes 4.1 actually good at tool calling look how fast this thing goes, video not sped up
0:12

someone submitted a pr that added "beast prompt" to opencode for gpt models it makes 4.1 actually good at tool calling look how fast this thing goes, video not sped up

dax

80,952 views • 11 months ago

Qwen3-Coder is now the default model powering Anycoder! Massive boost to productivity and creativity.
0:12

Qwen3-Coder is now the default model powering Anycoder! Massive boost to productivity and creativity.

Qwen

55,514 views • 10 months ago

Is this video sped up too fast?
0:12

Is this video sped up too fast?

She’s BigBoobs

12,353 views • 1 month ago

BREAKING: AI just replaced your coding assistant. Meet Qwen3‑Coder by Alibaba Group — it plans, codes, debugs & ships software on its own. I asked it to build a 2D space survival game. One prompt → working game → live in minutes.
0:16

BREAKING: AI just replaced your coding assistant. Meet Qwen3‑Coder by Alibaba Group — it plans, codes, debugs & ships software on its own. I asked it to build a 2D space survival game. One prompt → working game → live in minutes.

Dhaval Makwana

40,895 views • 10 months ago

It worked! Hermes Agent + Exo + Qwen3 Coder Next 8bit to create an incredible snake game, with model following 100% specifications passed in prompt! Let's load something bigger now!
0:15

It worked! Hermes Agent + Exo + Qwen3 Coder Next 8bit to create an incredible snake game, with model following 100% specifications passed in prompt! Let's load something bigger now!

Ivan Fioravanti ᯅ

159,666 views • 3 months ago

🔥Apple MLX first 6bit model is on Hugging Face!🔥 Qwen2.5-Coder-32B-Instruct-6bit! 3bit conversion and test in progress! Video 8x below on M4 Max 40GPU: - Prompt: 38 tokens, 61.731 tokens-per-sec - Generation: 1181 tokens, 16.939 tokens-per-sec - Peak memory: 25.122 GB
0:12

🔥Apple MLX first 6bit model is on Hugging Face!🔥 Qwen2.5-Coder-32B-Instruct-6bit! 3bit conversion and test in progress! Video 8x below on M4 Max 40GPU: - Prompt: 38 tokens, 61.731 tokens-per-sec - Generation: 1181 tokens, 16.939 tokens-per-sec - Peak memory: 25.122 GB

Ivan Fioravanti ᯅ

46,494 views • 1 year ago

in opencode v0.3.53 we implemented message queuing so you can send a message while opencode is working and it'll update the LLM on the next opportunity - basically lets you resteer it in a different direction should feel pretty natural
0:17

in opencode v0.3.53 we implemented message queuing so you can send a message while opencode is working and it'll update the LLM on the next opportunity - basically lets you resteer it in a different direction should feel pretty natural

dax

35,773 views • 10 months ago

Lio Rush is a FREAK OF NATURE. Absolutely tremendous wrestler. (No, this video is NOT sped up. Lio is actually this freaking fast.)
0:14

Lio Rush is a FREAK OF NATURE. Absolutely tremendous wrestler. (No, this video is NOT sped up. Lio is actually this freaking fast.)

Drainmaker

289,965 views • 1 year ago

MLX + M3 Ultra 512GB + Qwen3-235B-A22B-8bit = 🔥 "write a beautiful p5js particles animation that reacts to mouse clicks movements" Prompt: 22 tokens, 64.471 tokens-per-sec Generation: 6197 tokens, 18.916 tokens-per-sec Peak memory: 251.077 GB
0:14

MLX + M3 Ultra 512GB + Qwen3-235B-A22B-8bit = 🔥 "write a beautiful p5js particles animation that reacts to mouse clicks movements" Prompt: 22 tokens, 64.471 tokens-per-sec Generation: 6197 tokens, 18.916 tokens-per-sec Peak memory: 251.077 GB

Ivan Fioravanti ᯅ

24,160 views • 1 year ago

Introducing SHAI, OVHcloud’s Open Source, terminal-based AI coding assistant! ✅ OSS alternative to Codex/Claude ✅ Works out of the box w/ OVHcloud AI Endpoints (Works best with GPT-OSS or Qwen3 coder) ✅ MCP natively 🛠 Hacktoberfest contributions welcome! 🎉
0:19

Introducing SHAI, OVHcloud’s Open Source, terminal-based AI coding assistant! ✅ OSS alternative to Codex/Claude ✅ Works out of the box w/ OVHcloud AI Endpoints (Works best with GPT-OSS or Qwen3 coder) ✅ MCP natively 🛠 Hacktoberfest contributions welcome! 🎉

OVHcloud

20,458 views • 8 months ago

This video is not sped up
0:15

This video is not sped up

Jonathan | Fabworks 🏭

56,824 views • 11 months ago

This is not sped up. 😳
0:18

This is not sped up. 😳

WWE

913,116 views • 7 months ago

The video is not sped up.
0:18

The video is not sped up.

Johnny

11,115 views • 1 year ago

Seven Canyons Employee of the Month…every month! **this video is not sped up** he really shovels this fast all the time💪🏼
0:10

Sensitive content

Seven Canyons Employee of the Month…every month! **this video is not sped up** he really shovels this fast all the time💪🏼

Em Casey

28,908 views • 1 year ago

This is on M3 Ultra; M5 Ultra will be crazy!
0:16

This is on M3 Ultra; M5 Ultra will be crazy!

Ivan Fioravanti ᯅ

98,227 views • 7 months ago

Playing with Qwen3-TTS is and MLX-Audio locally on Mac Studio M3 Ultra 🔥 Amazing model by Qwen and great work by Prince Canuma bringing this magic to MLX! Tuning the voice with instructions feels like magic! Command and prompts to run it are in the video.
0:17

Playing with Qwen3-TTS is and MLX-Audio locally on Mac Studio M3 Ultra 🔥 Amazing model by Qwen and great work by Prince Canuma bringing this magic to MLX! Tuning the voice with instructions feels like magic! Command and prompts to run it are in the video.

Ivan Fioravanti ᯅ

35,720 views • 4 months ago

is this seriously not sped up LMAO
0:11

is this seriously not sped up LMAO

玥 is watching 有港回信 💌

1,344,497 views • 6 months ago

We promise this is NOT sped up 🏃‍♂️
0:19

We promise this is NOT sped up 🏃‍♂️

WWE

250,463 views • 7 months ago

opening files in Obsidian, not sped up — ez
0:11

opening files in Obsidian, not sped up — ez

kepano

948,318 views • 1 year ago