Loading video...

Video Failed to Load

Go Home

Running four high-level OpenCode agents + subagents with mlx_lm.server continuous batching and MiniMax M2.5 (6-bit). Fits easily on a 512GB M3 Ultra. Generation is quite fast. But prefill is still slow compared to cloud servers.

awnihannun's profile picture

Awni Hannun

38,393 subscribers

25,535 views • 3 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

Running four simultaneous OpenCode agents works well with mlx_lm.server continuous batching and MiniMax M2.1 on an M3 Ultra:
0:40

Running four simultaneous OpenCode agents works well with mlx_lm.server continuous batching and MiniMax M2.1 on an M3 Ultra:

Awni Hannun

95,013 views • 5 months ago

Running Minimax M2.1 (MiniMax (official)) with OpenCode (OpenCode) and mlx_lm.server. Works quite well on an M3 Ultra. Once the KV cache is warm the prompt processing is pretty quick. And token generation is very fast.
0:41

Running Minimax M2.1 (MiniMax (official)) with OpenCode (OpenCode) and mlx_lm.server. Works quite well on an M3 Ultra. Once the KV cache is warm the prompt processing is pretty quick. And token generation is very fast.

Awni Hannun

32,329 views • 5 months ago

I did it! It works! Using GLM-4.7-4bit with mlx_lm.server and opencode to fix real code locally! 🔥 Here single M3 Ultra 512GB, nex step phase will be 2 using Tensor Parallelism and then apply same changes to exo. Prefill is slow on a single machine, but generation is good.
1:44

I did it! It works! Using GLM-4.7-4bit with mlx_lm.server and opencode to fix real code locally! 🔥 Here single M3 Ultra 512GB, nex step phase will be 2 using Tensor Parallelism and then apply same changes to exo. Prefill is slow on a single machine, but generation is good.

Ivan Fioravanti ᯅ

44,000 views • 5 months ago

This is MiniMax-M2.5 MLX running in LM Studio on an Apple Mac Studio M3 Ultra 512GB. Fast enough out of the box for hosting OpenClaw, n8n workflows, and Open WebUI for the team.
0:53

This is MiniMax-M2.5 MLX running in LM Studio on an Apple Mac Studio M3 Ultra 512GB. Fast enough out of the box for hosting OpenClaw, n8n workflows, and Open WebUI for the team.

Patrick J Kennedy

73,547 views • 3 months ago

I didn't expect DeepSeek v4 PRO (not Flash) to run well on the Mac Studio M3 Ultra with 512GB of RAM. This is 2 bit quantized with the same DwarfStar recipe used for Flash. 433GB GGUF file. 130 t/s prefill, 13 t/s generation. Prefill in the video is low because small prompt.
1:24

I didn't expect DeepSeek v4 PRO (not Flash) to run well on the Mac Studio M3 Ultra with 512GB of RAM. This is 2 bit quantized with the same DwarfStar recipe used for Flash. 433GB GGUF file. 130 t/s prefill, 13 t/s generation. Prefill in the video is low because small prompt.

antirez

168,072 views • 19 days ago

Ollama now supports subagents and web search in Claude Code! Subagents can run tasks in parallel, such as file search, code exploration, and research, each in their own context. No MCP servers to configure or API keys required. Try it with any model on Ollama's cloud: ollama launch claude --model minimax-m2.5:cloud
0:43

Ollama now supports subagents and web search in Claude Code! Subagents can run tasks in parallel, such as file search, code exploration, and research, each in their own context. No MCP servers to configure or API keys required. Try it with any model on Ollama's cloud: ollama launch claude --model minimax-m2.5:cloud

ollama

83,979 views • 3 months ago

A perfect coding model for MLX on Apple silicon.. Qwen delivered again. Runs quite fast on an M3 Ultra. Running the 4-bit quantized with mlx-lm:
0:50

A perfect coding model for MLX on Apple silicon.. Qwen delivered again. Runs quite fast on an M3 Ultra. Running the 4-bit quantized with mlx-lm:

Awni Hannun

186,641 views • 10 months ago

PSA: MiniMax (official) M2.5 is freely available on OpenCode Great opportunity to check out the power of open models for coding
0:23

PSA: MiniMax (official) M2.5 is freely available on OpenCode Great opportunity to check out the power of open models for coding

Niels Rogge

49,016 views • 3 months ago

A long time coming but new mlx-lm is here with better batching support in the server and Gemma 4. pip install -U mlx-lm Here is a video where a single M3 Ultra serves 5 opencode sessions with Gemma 4 26B that process ~130k tokens in ~1.5 minutes.
2:03

A long time coming but new mlx-lm is here with better batching support in the server and Gemma 4. pip install -U mlx-lm Here is a video where a single M3 Ultra serves 5 opencode sessions with Gemma 4 26B that process ~130k tokens in ~1.5 minutes.

Angelos Katharopoulos

65,949 views • 1 month ago

GLM-4.7-8bit (350GB) running at 19 toks/s on two M3 Ultra 512GB using Tensor Parallelism with EXO - MLX, versus 14 toks/s with single node. 🚀 Now context benchmarking & then OpenCode tests 🔥 Note: this is from sources, I had to change things to run it.
0:50

GLM-4.7-8bit (350GB) running at 19 toks/s on two M3 Ultra 512GB using Tensor Parallelism with EXO - MLX, versus 14 toks/s with single node. 🚀 Now context benchmarking & then OpenCode tests 🔥 Note: this is from sources, I had to change things to run it.

Ivan Fioravanti ᯅ

327,687 views • 5 months ago

The new Deep Seek V3 0324 in 4-bit runs at > 20 toks/sec on a 512GB M3 Ultra with mlx-lm!
0:31

The new Deep Seek V3 0324 in 4-bit runs at > 20 toks/sec on a 512GB M3 Ultra with mlx-lm!

Awni Hannun

168,815 views • 1 year ago

Subagents running locally and simultaneously on MacBook Pro M5 with Codex CLI + LM Studio to review code and find bugs using Qwen 3.6 Powered by the updated MLX engine with batching in beta in the app The batching speed boost is noticeable
1:40

Subagents running locally and simultaneously on MacBook Pro M5 with Codex CLI + LM Studio to review code and find bugs using Qwen 3.6 Powered by the updated MLX engine with batching in beta in the app The batching speed boost is noticeable

Adrien Grondin

72,719 views • 15 days ago

MiniMax M2.5 > Claude Opus 4.6 I have been using it for a couple of days and have never hit the limit. I am currently on the Coding Plus plan which gives me 300 prompts per 5 hours. My new setup is OpenCode + MiniMax M2.5 What I have noticed: M2.5 – handles multi-step coding workflows and repo-level reasoning more reliably – better at chaining actions. think: generate test, fix bug, refactor, repeat – high TPS throughput M2.5 hits ~80%+ on SWE-Bench coding benchmarks on par with Opus and runs these tasks 37% faster than its predecessor while costing ~1/10th - 1/20th the price. Check out the video.
3:49

MiniMax M2.5 > Claude Opus 4.6 I have been using it for a couple of days and have never hit the limit. I am currently on the Coding Plus plan which gives me 300 prompts per 5 hours. My new setup is OpenCode + MiniMax M2.5 What I have noticed: M2.5 – handles multi-step coding workflows and repo-level reasoning more reliably – better at chaining actions. think: generate test, fix bug, refactor, repeat – high TPS throughput M2.5 hits ~80%+ on SWE-Bench coding benchmarks on par with Opus and runs these tasks 37% faster than its predecessor while costing ~1/10th - 1/20th the price. Check out the video.

Pratham

184,517 views • 3 months ago

🔥DeepSeek-v3-0324-8bit (~700GB) running on 2 x M3 Ultra 512GB ~12 tokens/sec Here the video!
0:25

🔥DeepSeek-v3-0324-8bit (~700GB) running on 2 x M3 Ultra 512GB ~12 tokens/sec Here the video!

Ivan Fioravanti ᯅ

19,761 views • 1 year ago

Opus 4.6 vs. Minimax M2.5 Prompt: Build an interactive solar system from scratch. Opus 4.6 tried to create a beautiful UI, but the sun’s shadow ruined it. Minimax M2.5 built a simple version that works beautifully. The winner is: 🥇 Minimax M2.5 🥈 Opus 4.6
0:49

Opus 4.6 vs. Minimax M2.5 Prompt: Build an interactive solar system from scratch. Opus 4.6 tried to create a beautiful UI, but the sun’s shadow ruined it. Minimax M2.5 built a simple version that works beautifully. The winner is: 🥇 Minimax M2.5 🥈 Opus 4.6

Okara

80,546 views • 3 months ago

The new Kimi K2 1T model (4-bit quant) runs on 2 512GB M3 Ultras with mlx-lm and mx.distributed. 1 trillion params, at a speed that's actually quite usable:
0:37

The new Kimi K2 1T model (4-bit quant) runs on 2 512GB M3 Ultras with mlx-lm and mx.distributed. 1 trillion params, at a speed that's actually quite usable:

Awni Hannun

237,964 views • 10 months ago

MiniMax M2.1 in 4-bit cruises on an M3 Ultra with mlx-lm. Generated a space invaders game using 5098 tokens at 47.2 tok/sec:
0:23

MiniMax M2.1 in 4-bit cruises on an M3 Ultra with mlx-lm. Generated a space invaders game using 5098 tokens at 47.2 tok/sec:

Awni Hannun

93,826 views • 5 months ago

Running GLM-4.7-Flash with OpenCode locally on M4 Max MacBook Pro. 4-bit model runs at 82 tok/sec. Prefill will get ~4x faster with M5 Max MacBook Pro (~28 Jan). EXO will also support disaggregating prefill and decode across devices, e.g. DGX Spark.
0:43

Running GLM-4.7-Flash with OpenCode locally on M4 Max MacBook Pro. 4-bit model runs at 82 tok/sec. Prefill will get ~4x faster with M5 Max MacBook Pro (~28 Jan). EXO will also support disaggregating prefill and decode across devices, e.g. DGX Spark.

Alex Cheema

127,296 views • 4 months ago

Minimax M3 is excellent at SVG generation, reaching close to Gemini 3.5 Flash levels and beating Opus 4.7 on SVG-Bench. With 1M context, native multimodality, strong agentic/coding ability and open weights coming soon, the closed-source moat is thinning fast. Full Video:
0:22

Minimax M3 is excellent at SVG generation, reaching close to Gemini 3.5 Flash levels and beating Opus 4.7 on SVG-Bench. With 1M context, native multimodality, strong agentic/coding ability and open weights coming soon, the closed-source moat is thinning fast. Full Video:

WorldofAI

16,499 views • 4 days ago

Compared MiniMax M3 against three other frontier open-source models: DeepSeek V4, Qwen 3.7 Max, and MiMo 2.5 Pro MiniMax M3 crashed others on: 1. collision physics 2. 3D pinball generation It was the only model that produced a result for the 3D pinball test For human motion rendering and hexagon sketching, Qwen 3.7 Max still tops
0:32

Compared MiniMax M3 against three other frontier open-source models: DeepSeek V4, Qwen 3.7 Max, and MiMo 2.5 Pro MiniMax M3 crashed others on: 1. collision physics 2. 3D pinball generation It was the only model that produced a result for the 3D pinball test For human motion rendering and hexagon sketching, Qwen 3.7 Max still tops

GMI Cloud

42,371 views • 3 days ago