Loading video...

Video Failed to Load

Go Home

GLM 5.2 has just been released 🔥 Here it's already running with MLX on two Mac Studios (M3 Ultra). This is comparable to the latest closed models, with weights you can download, quantize, distill, fine-tune, run.

pcuenq's profile picture

Pedro Cuenca

7,542 subscribers

83,438 views • 4 months ago •via X (Twitter)

EducationNews & PoliticsScience & Technology
LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

GLM-4.7 Flash by Z.ai running on M3 Ultra using MLX! 4bit at 81 toks/s 🔥 8bit at 64 toks/s 🔥 (video below) Model converted from transformers to MLX on the fly with Codex + GPT-5.2 HIgh!
0:19

GLM-4.7 Flash by Z.ai running on M3 Ultra using MLX! 4bit at 81 toks/s 🔥 8bit at 64 toks/s 🔥 (video below) Model converted from transformers to MLX on the fly with Codex + GPT-5.2 HIgh!

Ivan Fioravanti ᯅ

75,425 views • 9 months ago

Running Kimi K2.5 on my desk. Runs at 24 tok/sec with 2 x 512GB M3 Ultra Mac Studios connected with Thunderbolt 5 (RDMA) using EXO Labs / MLX backend. Yes, it can run clawdbot.
0:18

Running Kimi K2.5 on my desk. Runs at 24 tok/sec with 2 x 512GB M3 Ultra Mac Studios connected with Thunderbolt 5 (RDMA) using EXO Labs / MLX backend. Yes, it can run clawdbot.

Alex Cheema

3,319,962 views • 9 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 • 9 months ago

Whoa, GLM-5.2 is INSANE for UI/UX with the right prompting. Open models have finally closed the gap. You have to push it, but it's super creative. I just asked for a beautiful personal site; this was all the model's idea (multiple turns):
0:16

Whoa, GLM-5.2 is INSANE for UI/UX with the right prompting. Open models have finally closed the gap. You have to push it, but it's super creative. I just asked for a beautiful personal site; this was all the model's idea (multiple turns):

Anshu

178,213 views • 4 months ago

Okay this is kinda wild 👀 NVIDIA is basically handing out FREE API keys to 100+ top AI models - GLM 5.2, DeepSeek V4, Kimi K2.6, MiniMax M3, their own Nemotron and a ton more. it's called NVIDIA NIM and I had no idea it existed. it's rate-limited, so not something you'd run in production - but for personal use it's honestly great. you can poke a dozen frontier models, and some you can even grab and self-host. dropped a key into OpenCode and it just worked. the hyped ones like GLM 5.2 are jammed with queue right now, but the lighter models fly. all of it free.
0:15

Okay this is kinda wild 👀 NVIDIA is basically handing out FREE API keys to 100+ top AI models - GLM 5.2, DeepSeek V4, Kimi K2.6, MiniMax M3, their own Nemotron and a ton more. it's called NVIDIA NIM and I had no idea it existed. it's rate-limited, so not something you'd run in production - but for personal use it's honestly great. you can poke a dozen frontier models, and some you can even grab and self-host. dropped a key into OpenCode and it just worked. the hyped ones like GLM 5.2 are jammed with queue right now, but the lighter models fly. all of it free.

Stefan 3D AI

26,272 views • 3 months ago

MiniMax M2.5 casually responding at ~50 tok/s with MLX (M3 Ultra). The model was released one hour ago 🥳
0:14

MiniMax M2.5 casually responding at ~50 tok/s with MLX (M3 Ultra). The model was released one hour ago 🥳

Pedro Cuenca

170,567 views • 8 months ago

Running DeepSeek R1 on my desk Uses EXO Labs with Thunderbolt 5 interconnect (80Gbps) to run the full (671B, 8-bit) DeepSeek R1 distributed across 2 M3 Ultra 512GB Mac Studios (1TB total Unified Memory). Runs at 11 tok/sec. Theoretical max is ~20 tok/sec.
0:14

Running DeepSeek R1 on my desk Uses EXO Labs with Thunderbolt 5 interconnect (80Gbps) to run the full (671B, 8-bit) DeepSeek R1 distributed across 2 M3 Ultra 512GB Mac Studios (1TB total Unified Memory). Runs at 11 tok/sec. Theoretical max is ~20 tok/sec.

Alex Cheema

993,329 views • 1 year ago

made this game with GLM-5.2 (claude code) running on my 32x RTX 4090s
0:20

made this game with GLM-5.2 (claude code) running on my 32x RTX 4090s

Ning

26,465 views • 4 months ago

LLaVA Llama-3 and Phi-3 now on MLX 🎉🚀 You can now run inference locally on your Mac. pip install -U mlx-vlm I’m getting ~50 tokens on a M3 Max. Model cards 👇🏾
0:10

LLaVA Llama-3 and Phi-3 now on MLX 🎉🚀 You can now run inference locally on your Mac. pip install -U mlx-vlm I’m getting ~50 tokens on a M3 Max. Model cards 👇🏾

Prince Canuma

74,448 views • 2 years ago

🚨 do you understand what is happening?! OpenAI and Anthropic are putting the best models behind approval lists. Meanwhile GLM 5.2 is running locally on a Mac Studio, coding in a loop 24/7. If someone can turn off your AI, you are renting your intelligence layer. Local models are getting so good now that you can create gameplays like this.👇
0:18

🚨 do you understand what is happening?! OpenAI and Anthropic are putting the best models behind approval lists. Meanwhile GLM 5.2 is running locally on a Mac Studio, coding in a loop 24/7. If someone can turn off your AI, you are renting your intelligence layer. Local models are getting so good now that you can create gameplays like this.👇

Vadim

14,609 views • 4 months ago

🔴 GLM-5.5 could be just a few days away! Statistically, Zhipu has been releasing new models every 61 days on average this year GLM-4.7 → GLM-5: 51 days GLM-5 → GLM-5.1: 55 days GLM-5.1 → GLM-5.2: 70 days GLM-5.2 → GLM-5.3: 59 days Avg: 61 days Now: 76 days since GLM-5.3 Next GLM-5.5 est. ~Nov 3 (80%: Oct 30 – Nov 12)
0:19

🔴 GLM-5.5 could be just a few days away! Statistically, Zhipu has been releasing new models every 61 days on average this year GLM-4.7 → GLM-5: 51 days GLM-5 → GLM-5.1: 55 days GLM-5.1 → GLM-5.2: 70 days GLM-5.2 → GLM-5.3: 59 days Avg: 61 days Now: 76 days since GLM-5.3 Next GLM-5.5 est. ~Nov 3 (80%: Oct 30 – Nov 12)

Iam_

13,397 views • 7 days ago

Here's an example of a creation by GLM 5.2 on Design Arena, made from only one prompt. Just six months ago, there was a big enough difference between open and closed weight models that we considered having two separate evaluations. Open weights have come a long way. Congrats to the Z.ai team for this achievement!
0:17

Here's an example of a creation by GLM 5.2 on Design Arena, made from only one prompt. Just six months ago, there was a big enough difference between open and closed weight models that we considered having two separate evaluations. Open weights have come a long way. Congrats to the Z.ai team for this achievement!

Grace Li

14,570 views • 4 months ago

LM Studio 0.3.4 ships with Apple MLX 🚢🍎 Run on-device LLMs super fast, 100% locally and offline on your Apple Silicon Mac! Includes: > run Llama 3.2 1B at ~250 tok/sec (!) on M3 > enforce structured JSON responses > use via chat UI, or from your own code > run multiple models simultaneously > download any model from Hugging Face Video at 1x speed.
0:19

LM Studio 0.3.4 ships with Apple MLX 🚢🍎 Run on-device LLMs super fast, 100% locally and offline on your Apple Silicon Mac! Includes: > run Llama 3.2 1B at ~250 tok/sec (!) on M3 > enforce structured JSON responses > use via chat UI, or from your own code > run multiple models simultaneously > download any model from Hugging Face Video at 1x speed.

LM Studio

171,873 views • 2 years ago

Sam 3 by Facebook now on MLX 🚀 Here is a realtime object tracking running on M3 Max 96GB.
0:16

Sam 3 by Facebook now on MLX 🚀 Here is a realtime object tracking running on M3 Max 96GB.

Prince Canuma

180,498 views • 7 months ago

Apple just showcased the ultimate setup for running AI models locally. These are 4 M5 Ultra Mac Studios running Kimi K2.7 Code locally to fix a bug in a 3D rendered scene on XCode. Each has 256GB RAM, so thats 1TB combined memory. Enough to run the biggest open AI models locally like Llama 4, Deepseek v3 and Qwen 3. Perfect for agentic coding when the project codebase is sensitive and you wan to keep it private. This is easily a $50K setup!
0:15

Apple just showcased the ultimate setup for running AI models locally. These are 4 M5 Ultra Mac Studios running Kimi K2.7 Code locally to fix a bug in a 3D rendered scene on XCode. Each has 256GB RAM, so thats 1TB combined memory. Enough to run the biggest open AI models locally like Llama 4, Deepseek v3 and Qwen 3. Perfect for agentic coding when the project codebase is sensitive and you wan to keep it private. This is easily a $50K setup!

Ishan Sharma

17,819 views • 1 month ago

Ready! Is there a brave Apple Silicon user with a 64GB machine willing to test DwarfStar with Qwen 3.8 Flash Next? In the video it runs at 65 t/s on M3 Ultra. 🚀 I uploaded Q2 weights here: PR branch is here:
0:17

Ready! Is there a brave Apple Silicon user with a 64GB machine willing to test DwarfStar with Qwen 3.8 Flash Next? In the video it runs at 65 t/s on M3 Ultra. 🚀 I uploaded Q2 weights here: PR branch is here:

Ivan Fioravanti

36,808 views • 2 months ago

Qwen 3 0.6B is Ultra powerful if fine-tuned! Here I got 83.5% on classification tasks! Only Gemini 2.5 Pro does better with 85%! Video is in real time on M3 Ultra!
0:19

Qwen 3 0.6B is Ultra powerful if fine-tuned! Here I got 83.5% on classification tasks! Only Gemini 2.5 Pro does better with 85%! Video is in real time on M3 Ultra!

Ivan Fioravanti ᯅ

143,143 views • 1 year ago

Tested Muse Spark 1.2, Kimi K3, and GLM 5.2 with our FlappyBench 3 models, same prompt with the /design command. Reviewed gameplay features, UX/UI, and cost. 🔹 Kimi K3 → 9.5/10 · $0.0740 🔹 GLM 5.2 → 9/10 · $0.0480 🔹 Muse Spark 1.2 → 8/10 · $0.0187 Muse Spark 1.2 delivered a completely different game style despite being super cheap. Kimi K3 and GLM 5.2 outputs are comparable and close in cost.
0:19

Tested Muse Spark 1.2, Kimi K3, and GLM 5.2 with our FlappyBench 3 models, same prompt with the /design command. Reviewed gameplay features, UX/UI, and cost. 🔹 Kimi K3 → 9.5/10 · $0.0740 🔹 GLM 5.2 → 9/10 · $0.0480 🔹 Muse Spark 1.2 → 8/10 · $0.0187 Muse Spark 1.2 delivered a completely different game style despite being super cheap. Kimi K3 and GLM 5.2 outputs are comparable and close in cost.

Command Code

23,657 views • 3 months ago