🎮 Tech, gaming, AI, and everything in between. 🤖 Building with it, not just talking about it. 🔥 From the mind of @ToNYD2WiLD
- Tweets 3,693
- Following 203
- Followers 4,259
- Likes 886
Pinned Tweet
Best Local AI Models RIGHT NOW! (1× RTX 3090, 4× RTX 3090, 1× DGX Spark & 4× DGX Spark) piped.video/poaxCDCmxbs
Made with AI
4
11
1,635
Stuff like this will only get worse as A.I. progresses. The only way to stay ahead is the ability to make and have NEW ideas.
We’re now in the copy the HF quantization repo and put your name on it as an expert phase of Local AI
That’s not how Opensource as a community flourishes
You might have good intentions, but please do right by others & lift them up with you rather than falsely crediting yourself
8
1,531
🎙️ Your DeepSeek Harness agent can now TALK BACK 🗣️🔥
🎤 Dictation that streams live
💬 Hands-free conversation mode
✋ Butt in & steer it mid-answer
🏠 All local, no API keys
🔗 github.com/tonyd2wild/DeepSe…
🔊 github.com/tonyd2wild/DeepSe…
🧰 github.com/tonyd2wild/DeepSe…
3
2
34
2,590
I got a bigger update coming !
🎙️ New drop for the DeepSeek Harness: push-to-dictate!
Click the mic, talk, click again → your words land in the chat ✍️
🔒 100% local: faster-whisper on your own machine. No cloud, no API key
⚡ ~2s per phrase on CPU
🔗 github.com/tonyd2wild/DeepSe…
🧰 github.com/tonyd2wild/DeepSe…
9
1,210
Going to look to just make this STREAM in real time instead of having to press the button twice also a auto send so you can have a REALTIME Chat back and forth effect.
🎙️ New drop for the DeepSeek Harness: push-to-dictate!
Click the mic, talk, click again → your words land in the chat ✍️
🔒 100% local: faster-whisper on your own machine. No cloud, no API key
⚡ ~2s per phrase on CPU
🔗 github.com/tonyd2wild/DeepSe…
🧰 github.com/tonyd2wild/DeepSe…
1
524
🎙️ New drop for the DeepSeek Harness: push-to-dictate!
Click the mic, talk, click again → your words land in the chat ✍️
🔒 100% local: faster-whisper on your own machine. No cloud, no API key
⚡ ~2s per phrase on CPU
🔗 github.com/tonyd2wild/DeepSe…
🧰 github.com/tonyd2wild/DeepSe…
4
1
28
3,229
Link to Repo - github.com/tonyd2wild/DeepSe…
2
194
Tech2Wild retweeted
People will HATE A.I. and say it is taking jobs and they rather pay a human to do the work. Not knowing the human they are paying to do the work is using A.I.
1
3
6
1,523
People in the sneaker world HATE A.I. I always have to have polar opposite audience bases...
1
6
807
I think Mimo V3 is going to put Mimo BACK in the game with the BIG BOYS. Mimo V2.6 Flash is decent but it is more of a catch up. I do feel like Mimo V3 is coming in October that is why they released V2.6 without any build up.
1
5
1,149
So We Jumping From GLM 5.3 Flash to GLM 5.5 Flash ?! October is Looking VERY EXCITING !
🚨惊了！OpenCode 数据页提前曝光下一批模型，目录已经挂上！
这不是官宣能用，是模型 ID 已经进库：
🔹Kimi K4（Moonshot）
opencode.ai/data/moonshot/ki…
🔹GLM 5.5 Flash（Zhipu）
opencode.ai/data/zhipu/glm-5…
🔹DeepSeek V4.1 Pro
opencode.ai/data/deepseek/de…
🔹Qwen 3.8 Max Preview Free
opencode.ai/data/qwen/qwen3-…
🔹Muse Spark 1.4 Contributor（Meta）
opencode.ai/data/meta/muse-s…
🔹额外同批出现：
Qwen 3.8 Max Prime（列表标注 9/23，暂无用量）
opencode.ai/data/alibaba
现役还是 K3 / GLM-5.3-Flash / V4.1 Flash。K4、5.5、V4.1 Pro 才是下一代信号。
#OpenCode #AI #KimiK4 #GLM55 #DeepSeek #Qwen #MuseSpark #OpenSource
5
1
32
3,242
Even though GLM 5.3 Flash is my GO - TO... These numbers are tempting and impressive...
More DS4.1 TP4 improvements on 4x DGX Spark 🚀
- Prose decode: 85 tok/s at c1 (was 69)
- Boot: ~3 min
- Prefill: ~4.8k tok/s at 32k–128k, 4.3k at 262k
- KV cache: ~6.5M tokens, 1M context
- Same weights, same quality (qeval unchanged)
github.com/knapcio/DeepSeek-…
2
1
13
3,047
That could be possible but Deepseek has been a very verbose model. It does alot of talk in place. GLM has been solid. I used both, both work. I went back to GLM this week and it just seems to be a little more polished.
8
1
28
4,105
My pick for the best model on DGX Spark, as of today 👇
1️⃣ 1 Spark: Qwen 3.8 Flash
2️⃣ 2 Sparks: GLM 5.3 Flash (NVFP4 + DFlash2, 262K context)
3️⃣ 3 Sparks: GLM 5.3 Flash
4️⃣ 4 Sparks: GLM 5.3 Flash (500K context)
You could run the full GLM 5.3, but for 2+ Sparks, Flash is still the one I'd run today 🔥
TP2 recipe 👇
github.com/tonyd2wild/GLM-5.…
14
5
100
8,247
