# ![@victormustar Avatar](https://lunarcrush.com/gi/w:26/cr:twitter::784830007.png) @victormustar Victor M

Victor M posts on X about ai, inference, llamacpp, claude code the most. They currently have [------] followers and [---] posts still getting attention that total [-------] engagements in the last [--] hours.

### Engagements: [-------] [#](/creator/twitter::784830007/interactions)
![Engagements Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::784830007/c:line/m:interactions.svg)

- [--] Week [-------] +47%
- [--] Month [---------] +132%
- [--] Months [----------] +14%
- [--] Year [----------] +563%

### Mentions: [--] [#](/creator/twitter::784830007/posts_active)
![Mentions Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::784830007/c:line/m:posts_active.svg)

- [--] Week [--] -51%
- [--] Month [--] -1.20%
- [--] Months [---] +179%
- [--] Year [---] +162%

### Followers: [------] [#](/creator/twitter::784830007/followers)
![Followers Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::784830007/c:line/m:followers.svg)

- [--] Week [------] +0.37%
- [--] Month [------] +5%
- [--] Months [------] +25%
- [--] Year [------] +126%

### CreatorRank: [------] [#](/creator/twitter::784830007/influencer_rank)
![CreatorRank Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::784830007/c:line/m:influencer_rank.svg)

### Social Influence

**Social category influence**
[technology brands](/list/technology-brands)  17.65% [stocks](/list/stocks)  #758 [gaming](/list/gaming)  0.98%

**Social topic influence**
[ai](/topic/ai) #1368, [inference](/topic/inference) #5, [llamacpp](/topic/llamacpp) 6.86%, [claude code](/topic/claude-code) 3.92%, [microsoft](/topic/microsoft) 3.92%, [$9month](/topic/$9month) #1, [pi](/topic/pi) #1327, [if you](/topic/if-you) 2.94%, [meta](/topic/meta) 1.96%, [agentic](/topic/agentic) 1.96%

**Top accounts mentioned or mentioned by**
[@maziyarpanahi](/creator/undefined) [@alexellisuk](/creator/undefined) [@benburtenshaw](/creator/undefined) [@patelnav](/creator/undefined) [@huggingface](/creator/undefined) [@akhaliq](/creator/undefined) [@cocktailpeanut](/creator/undefined) [@tommyfalkowski](/creator/undefined) [@zaiorg](/creator/undefined) [@direitolibertr1](/creator/undefined) [@airbnb](/creator/undefined) [@paradite](/creator/undefined) [@mistralai](/creator/undefined) [@mistralaidevs](/creator/undefined) [@ngxson](/creator/undefined) [@flowvsgravity](/creator/undefined) [@haareblond](/creator/undefined) [@openclaw](/creator/undefined) [@antirez](/creator/undefined) [@thinkfulearn](/creator/undefined)

**Top assets mentioned**
[Microsoft Corp. (MSFT)](/topic/microsoft) [Alphabet Inc Class A (GOOGL)](/topic/$googl)
### Top Social Posts
Top posts by engagements in the last [--] hours

"The Secrets feature is quite interesting Introducing Deno Sandbox: Instant Linux microVMs TypeScript and Python () SDKs + REST API Secure against prompt injection attacks https://t.co/E4r4oefDrm Introducing Deno Sandbox: Instant Linux microVMs TypeScript and Python () SDKs + REST API Secure against prompt injection attacks https://t.co/E4r4oefDrm"  
[X Link](https://x.com/victormustar/status/2018773726092038232)  2026-02-03T19:48Z 23.3K followers, 101.2K engagements


"🤯 Is 3D generation finally cracked TRELLIS might be the solution we've been waiting for nearly ready for production Plus theres a free ZeroGPU demo available for everyone to try out"  
[X Link](https://x.com/victormustar/status/1864993373964156985)  2024-12-06T11:20Z 23.2K followers, 153.2K engagements


"Ok let's be honest you really cooked with this one Google 😗"  
[X Link](https://x.com/victormustar/status/1900291486115127420)  2025-03-13T21:02Z 23.2K followers, 1.6M engagements


"Crazy what you can do on Hugging Face in [--] seconds rejoiced by the rebirth of the skeuomorphic isomorphic 3D icons on @Airbnb I've trained a FLUX LoRA to generate 3D icons in that style https://t.co/XoipU9evNA rejoiced by the rebirth of the skeuomorphic isomorphic 3D icons on @Airbnb I've trained a FLUX LoRA to generate 3D icons in that style https://t.co/XoipU9evNA"  
[X Link](https://x.com/victormustar/status/1922944620314390855)  2025-05-15T09:18Z 23.2K followers, 447.5K engagements


"Blown away by OmniSVG. Bitmap to vector is looking crazy clean vectorization breakthrough might actually be here 👀🚀 👇Demo available on Hugging Face"  
[X Link](https://x.com/victormustar/status/1949936038996037894)  2025-07-28T20:52Z 23.2K followers, 57.4K engagements


"The before/after on GLM-4.7-Flash for local agents is crazy. Less overthinking reliable tool calling it just works 🔥 You can now run GLM-4.7-Flash locally on your device🔥 GLM-4.7-Flash is the best performing 30B model on SWE-Bench and GPQA. With 200K context it excels at coding agents chat & reasoning. Run local with 24GB RAM. Guide: https://t.co/SpJxl00VIa GGUF: https://t.co/aTuUxu32z3 https://t.co/3MwNRe3iva You can now run GLM-4.7-Flash locally on your device🔥 GLM-4.7-Flash is the best performing 30B model on SWE-Bench and GPQA. With 200K context it excels at coding agents chat &"  
[X Link](https://x.com/victormustar/status/2013550336703672593)  2026-01-20T09:53Z 23.2K followers, 25.1K engagements


"OMG Z-Image dropped on Hugging Face 🤯 https://huggingface.co/Tongyi-MAI/Z-Image https://huggingface.co/Tongyi-MAI/Z-Image"  
[X Link](https://x.com/victormustar/status/2016185392332943470)  2026-01-27T16:23Z 23.2K followers, 16.7K engagements


"Cannot believe KIMI K2.5 is that good. open source has caught up I think 🤯"  
[X Link](https://x.com/victormustar/status/2016446140145643575)  2026-01-28T09:39Z 23.2K followers, [----] engagements


"who is running Kimi K2.5 locally (and so can survive the apocalypse) https://huggingface.co/posts/danielhanchen/602916263790271 https://huggingface.co/posts/danielhanchen/602916263790271"  
[X Link](https://x.com/victormustar/status/2016528437855269087)  2026-01-28T15:06Z 23.2K followers, [----] engagements


"wait Claude Code doesn't support 😬 http://AGENTS.md http://AGENTS.md"  
[X Link](https://x.com/victormustar/status/2016852941492121612)  2026-01-29T12:36Z 23.2K followers, 19.5K engagements


"how many mac mini to run this 👀 https://huggingface.co/mlx-community/Kimi-K2.5 https://huggingface.co/mlx-community/Kimi-K2.5"  
[X Link](https://x.com/victormustar/status/2016936349522153546)  2026-01-29T18:07Z 23.2K followers, 15.3K engagements


"back to when nobody cared about AI 🥹"  
[X Link](https://x.com/victormustar/status/2017232469792092229)  2026-01-30T13:44Z 23.3K followers, [----] engagements


"exports in glb is quite nice on this one Huge drop by Meta: ActionMesh turns any video into an animated 3D mesh. Demo available on Hugging Face https://t.co/dDh144uLuP Huge drop by Meta: ActionMesh turns any video into an animated 3D mesh. Demo available on Hugging Face https://t.co/dDh144uLuP"  
[X Link](https://x.com/victormustar/status/2017276433169555698)  2026-01-30T16:39Z 23.2K followers, [----] engagements


"that's not wrong GPT-OSS 120b was dismissed too quickly (I don't know why) but it's a great model. we need a [----] update now is a great timing for OpenAI to drop it 👀"  
[X Link](https://x.com/victormustar/status/2017537967632331143)  2026-01-31T09:58Z 23.2K followers, 24.5K engagements


"I really want to start sponsoring people on HF - not sure we should implement the feature or not 👀"  
[X Link](https://x.com/victormustar/status/2018624024041906536)  2026-02-03T09:54Z 23.2K followers, [----] engagements


"Qwen delivered 🔥 absolute must try: [--] Run llama-server -hf unsloth/Qwen3-Coder-Next-GGUF:Q4_K_M 2"Make me a Counter strIKE GAME" 🚀 IntroducingQwen3-Coder-Next an open-weight LM built for coding agents & local development. Whats new: 🤖 Scaling agentic training:800K verifiable tasks + executable envs 📈 EfficiencyPerformance Tradeoff: achieves strong results on SWE-Bench Pro with 80B total params and https://t.co/P7BmZwdaQ9 🚀 IntroducingQwen3-Coder-Next an open-weight LM built for coding agents & local development. Whats new: 🤖 Scaling agentic training:800K verifiable tasks + executable"  
[X Link](https://x.com/victormustar/status/2018741278922649601)  2026-02-03T17:40Z 23.2K followers, 15.6K engagements


"I don't get the obsession with using Claude Code for open models. Just use an open source harness one where you can read the source and that isn't optimized for a particular model. My current stack is Pi + llama.cpp and I'm super happy with it: crazy.now you can run claude code locally. install claude code run it with ollama model 100% free and local. https://t.co/hf3FapbgK5 crazy.now you can run claude code locally. install claude code run it with ollama model 100% free and local. https://t.co/hf3FapbgK5"  
[X Link](https://x.com/victormustar/status/2019014809669271610)  2026-02-04T11:46Z 23.2K followers, 16.8K engagements


"@alexellisuk M4 Max 128GB"  
[X Link](https://x.com/victormustar/status/2019058706546086016)  2026-02-04T14:41Z 23.2K followers, [---] engagements


"@paradite_ yes it was not a great signal. anthropic branding = focused on important things not competitors"  
[X Link](https://x.com/victormustar/status/2019372505899888772)  2026-02-05T11:28Z 23.2K followers, [---] engagements


"GPT [---] Codex did great on my "detailed-Boeing-747-using-THREEJS-primitives" benchmark. (great at spatial awareness) Here's where it gets interesting: I also prompted it to enter a loop of taking screenshots from different angles and improving the Boeing until satisfied it worked much better than I expected it's seems models start to get good at long iterative loops on (semi)verifiable problems - before they were falling miserably at this 👀 My new favorite benchmark: 3D modeling a Boeing [---] using THREE.js. It's more about spatial understanding than library knowledge (few methods needed). So"  
[X Link](https://x.com/victormustar/status/2019529666269212769)  2026-02-05T21:52Z 23.3K followers, 21.7K engagements


"Great app: Voxtral-Subtitles to transcribe any video with word-level subtitles speaker diarization & multilingual translation. Try in now on Hugging Face Spaces"  
[X Link](https://x.com/victormustar/status/2020776196343238656)  2026-02-09T08:26Z 23.3K followers, 14.4K engagements


"Here: https://huggingface.co/spaces/pandora-s/Voxtral-Subtitles https://huggingface.co/spaces/pandora-s/Voxtral-Subtitles"  
[X Link](https://x.com/victormustar/status/2020776198650118183)  2026-02-09T08:26Z 23.3K followers, [----] engagements


"Pi now supports Hugging Face Inference 💫 If you have a HF account - try Kimi K-2.5: scary smart model running 5x faster than what you're used to. Honestly feels like the future of working with agents"  
[X Link](https://x.com/victormustar/status/2020919604416061781)  2026-02-09T17:55Z 23.3K followers, 13.8K engagements


"Get Pi then /models and type "Hugging" to view available models. https://pi.dev/ https://pi.dev/"  
[X Link](https://x.com/victormustar/status/2020919606647349294)  2026-02-09T17:55Z 23.3K followers, [----] engagements


"turning point time to reconnect with nature Oh my Tennis with Seedance [---] This is 100% AI generated. Let that sink in. https://t.co/bokwajI5b6 Oh my Tennis with Seedance [---] This is 100% AI generated. Let that sink in. https://t.co/bokwajI5b6"  
[X Link](https://x.com/victormustar/status/2021203403872010659)  2026-02-10T12:43Z 23.3K followers, [----] engagements


"Too much AI services turning out to be scams lately. Just get Hugging Face PRO for $9/month imo: Generate images with FLUX.2 Klein Z-Image Turbo Generate videos with Wan2.2 Animate LTX-2 Turbo CogVideoX-5B Generate 3D models with Microsoft TRELLIS.2 Tencent Hunyuan3D-2.1 Generate music with ACE-Step Clone voices with F5-TTS Chatterbox Qwen3-TTS Edit images with Omni Image Editor Qwen Image Edit Upscale images with Finegrain Image Enhancer Tile Upscaler Remove backgrounds with BRIA RMBG [---] BiRefNet Virtual try-on with IDM-VTON Lip sync videos with LatentSync OmniAvatar OCR documents with"  
[X Link](https://x.com/victormustar/status/2021631509413679576)  2026-02-11T17:04Z 23.3K followers, 55.4K engagements


"Voxtral Mini Realtime is exceptional at real-time transcription (I need this on my phone now.) Try the demo on Hugging Face Introducing Voxtral Transcribe [--] next-gen speech-to-text models by @MistralAI. State-of-the-art transcription speaker diarization sub-200ms real-time latency. Details in 🧵 https://t.co/0IeiJOpiAZ Introducing Voxtral Transcribe [--] next-gen speech-to-text models by @MistralAI. State-of-the-art transcription speaker diarization sub-200ms real-time latency. Details in 🧵 https://t.co/0IeiJOpiAZ"  
[X Link](https://x.com/victormustar/status/2019330353328963830)  2026-02-05T08:40Z 23.3K followers, 23.3K engagements


"let's go: https://huggingface.co/spaces/mistralai/Voxtral-Mini-Realtime https://huggingface.co/spaces/mistralai/Voxtral-Mini-Realtime"  
[X Link](https://x.com/victormustar/status/2019330355166117961)  2026-02-05T08:40Z 23.3K followers, [----] engagements


"RT @MistralAIDevs: Automate or simplify your workflow for subtitle creation with Voxtral Transcribe [---] From word-level timestamps to diar"  
[X Link](https://x.com/victormustar/status/2019520762517106779)  2026-02-05T21:17Z 23.3K followers, [--] engagements


"RT @ngxson: Qwen3-Coder-Next and Minimax-M2.1 are available on HF inference endpoints with the price of $2.5/hr and $5/hr respectively. Wi"  
[X Link](https://x.com/victormustar/status/2020921825471938790)  2026-02-09T18:04Z 23.3K followers, [--] engagements


"RT @ben_burtenshaw: this happened and it's totally normal"  
[X Link](https://x.com/victormustar/status/2020965241891537245)  2026-02-09T20:57Z 23.3K followers, [--] engagements


"@patelnav it should select the best one available on inference https://huggingface.co/inference/models https://huggingface.co/inference/models"  
[X Link](https://x.com/victormustar/status/2021294554008404034)  2026-02-10T18:45Z 23.3K followers, [--] engagements


"RT @MaziyarPanahi: Always nice to see trending on @huggingface"  
[X Link](https://x.com/victormustar/status/2021331470363607087)  2026-02-10T21:12Z 23.3K followers, [--] engagements


"RT @_akhaliq: Anthropic 🤝 Hugging Face"  
[X Link](https://x.com/victormustar/status/2021526630280364129)  2026-02-11T10:08Z 23.3K followers, [--] engagements


"Introducing: AI Video Composer 🔥 Drag and drop your assets (images/videos/audios) to create any video you want using natural language 100% Open Source and powered by Qwen2.5-Coder-32B (this thing is a beast) https://huggingface.co/spaces/huggingface-projects/ai-video-composer https://huggingface.co/spaces/huggingface-projects/ai-video-composer"  
[X Link](https://x.com/victormustar/status/1860744979884163417)  2024-11-24T17:59Z 23.3K followers, 266.6K engagements


"Grok [--] is now #1 trending on Hugging Face 💫"  
[X Link](https://x.com/victormustar/status/1960613514562752685)  2025-08-27T08:01Z 23.3K followers, 7.4M engagements


"🎉 llama.cpp now has Ollama-style model management. Auto-discover GGUFs from cache Load on first request Each model runs in its own process Route by model (OpenAI-compatible API) LRU unload at --models-max https://huggingface.co/blog/ggml-org/model-management-in-llamacpp https://huggingface.co/blog/ggml-org/model-management-in-llamacpp"  
[X Link](https://x.com/victormustar/status/1999484435910263256)  2025-12-12T14:20Z 23.3K followers, 35.9K engagements


"Funny how you can do the same thing with a local model without sending all your data to a remote cloud. In Cowork you give Claude access to a folder on your computer. Claude can then read edit or create files in that folder. Try it to create a spreadsheet from a pile of screenshots or produce a first draft from scattered notes. https://t.co/GEaMgDksUp In Cowork you give Claude access to a folder on your computer. Claude can then read edit or create files in that folder. Try it to create a spreadsheet from a pile of screenshots or produce a first draft from scattered notes."  
[X Link](https://x.com/victormustar/status/2011078287762825474)  2026-01-13T14:10Z 23.3K followers, 1.2M engagements


"Qwen3-Coder-Next is now on HuggingChat 😍 Try it now with your free HF credits (it's very good)"  
[X Link](https://x.com/victormustar/status/2018999276169465971)  2026-02-04T10:45Z 23.3K followers, 13K engagements


"RT @cocktailpeanut: PSA: Gradio is an API disguised as a UI. The UI is just one client"  
[X Link](https://x.com/victormustar/status/2021230675307675919)  2026-02-10T14:32Z 23.3K followers, [--] engagements


"My new favorite benchmark: 3D modeling a Boeing [---] using THREE.js. It's more about spatial understanding than library knowledge (few methods needed). So far no model gets it right 👀👀"  
[X Link](https://x.com/victormustar/status/1916573220863439296)  2025-04-27T19:20Z 23.3K followers, 78.2K engagements


"Hugging Face PRO is the wildest $9/month deal in AI right now🤯 🔹 [--] min/day of H200 compute on Spaces ZeroGPU 🔹 1M free inference tokens from 15+ providers (Groq Cerebras etc.) 🔹 1TB private storage and more 🔹 More cool things"  
[X Link](https://x.com/victormustar/status/1935285458394583356)  2025-06-18T10:36Z 23.3K followers, 728.8K engagements


"RT @_akhaliq: MiniCPM-o-4.5 streaming app is out on Hugging Face app: https://huggingface.co/spaces/openbmb/MiniCPM-o-4_5-Demo https://huggingface.co/spaces/openbmb/MiniCPM-o-4_5-Demo"  
[X Link](https://x.com/victormustar/status/2021628017563595058)  2026-02-11T16:50Z 23.3K followers, [--] engagements


"life feels weird since AI is working 👀"  
[X Link](https://x.com/victormustar/status/2021328371477581960)  2026-02-10T21:00Z 23.3K followers, [---] engagements


"GLM-5 made no obvious mistakes on my Boeing [---] using THREEJS benchmark. First open source model to do it My new favorite benchmark: 3D modeling a Boeing [---] using THREE.js. It's more about spatial understanding than library knowledge (few methods needed). So far no model gets it right 👀👀 https://t.co/op9Nj5ULP8 My new favorite benchmark: 3D modeling a Boeing [---] using THREE.js. It's more about spatial understanding than library knowledge (few methods needed). So far no model gets it right 👀👀 https://t.co/op9Nj5ULP8"  
[X Link](https://x.com/victormustar/status/2021578607634551075)  2026-02-11T13:34Z 23.3K followers, 46.2K engagements


"next level https://huggingface.co/zai-org/GLM-5 https://huggingface.co/zai-org/GLM-5"  
[X Link](https://x.com/victormustar/status/2021644170688135351)  2026-02-11T17:55Z 23.3K followers, [----] engagements


"follow https://huggingface.co/zai-org https://huggingface.co/zai-org"  
[X Link](https://x.com/victormustar/status/2021644260895044012)  2026-02-11T17:55Z 23.3K followers, [---] engagements


"Huge drop by Meta: ActionMesh turns any video into an animated 3D mesh. Demo available on Hugging Face"  
[X Link](https://x.com/victormustar/status/2017275263621435729)  2026-01-30T16:34Z 23.3K followers, 95.6K engagements


"RT @ben_burtenshaw: Eval scores in [----] are broken. MMLU at 91%+ GSM8K at 94%+ yet models still can't handle basic multi-step tasks. And"  
[X Link](https://x.com/victormustar/status/2019804290760724898)  2026-02-06T16:04Z 23.3K followers, [--] engagements


"is there a GUI to manage install/disable Skills and symlink between the agents apps because this is getting hard rn"  
[X Link](https://x.com/victormustar/status/2021183299670974489)  2026-02-10T11:23Z 23.3K followers, [----] engagements


"http://x.com/i/article/2020875588999786497 http://x.com/i/article/2020875588999786497"  
[X Link](https://x.com/victormustar/status/2021964315994046897)  2026-02-12T15:07Z 23.3K followers, [----] engagements


"DeepSeek-V3-0324 thank you for everything 🧡 Probably the best non reasoning and it's open source Use this app to vibe code with it (just made a SimCity)"  
[X Link](https://x.com/victormustar/status/1905700050308915578)  2025-03-28T19:14Z 23.2K followers, 46.8K engagements


"Breakthrough The new king of 3D gen And it's open source with a free online demo 🤯"  
[X Link](https://x.com/victormustar/status/1905914007451525144)  2025-03-29T09:24Z 23.2K followers, 86.9K engagements


"DeepSeek-V3-0324 is insane at UI 🤯 Things are getting wild someone prompted this 👀 Build your app now: (no login required) http://hf.co/spaces/enzostvs/deepsite http://hf.co/spaces/enzostvs/deepsite"  
[X Link](https://x.com/victormustar/status/1907404106043420788)  2025-04-02T12:05Z 23.2K followers, 169.6K engagements


"I think you forget that one: 100% open source UI + model (DeepSite) Three AIs one prompt. Who did it best 👇 I asked all three to 'make a Clockify clone'. https://t.co/XFm3lBDI2Q Three AIs one prompt. Who did it best 👇 I asked all three to 'make a Clockify clone'. https://t.co/XFm3lBDI2Q"  
[X Link](https://x.com/victormustar/status/1907500824734912664)  2025-04-02T18:30Z 23.2K followers, 43.8K engagements


"Insane Text-To-Texture dropped on Hugging Face 🤯 Try it here: http://hf.co/spaces/VAST-AI/MV-Adapter-Text2Texture http://hf.co/spaces/VAST-AI/MV-Adapter-Text2Texture"  
[X Link](https://x.com/victormustar/status/1909499328961589297)  2025-04-08T06:51Z 23.2K followers, 40K engagements


"👀 ByteDanceResearch just opensourced ChatTS14B natively understands & reasons over timeseries (Apache2.0 licensed) https://huggingface.co/bytedance-research/ChatTS-14B https://huggingface.co/bytedance-research/ChatTS-14B"  
[X Link](https://x.com/victormustar/status/1913558120799912363)  2025-04-19T11:39Z 23.2K followers, 70.9K engagements


"Dia is absolutely stunning 🤯 1.6B parameter TTS model to create realistic dialogue from text. Control emotion/tone via audio conditioning + generates nonverbals like laughter & coughs. Licensed Apache [---] 🔥 Sharing the online demo below"  
[X Link](https://x.com/victormustar/status/1914599099334095321)  2025-04-22T08:36Z 23.2K followers, 40.7K engagements


"Qwen/Qwen3-30B-A3B is absolute game changer 🤯 Are you on a Mac then use MLX weights - getting [---] tokens/sec here. It changes everything for real world use cases"  
[X Link](https://x.com/victormustar/status/1917144161377017862)  2025-04-29T09:09Z 23.2K followers, 144.2K engagements


"Your Mac can generate unlimited awesome speech audio 100% local in a single command thanks to Dia and MLX 🔥 Copy/paste this command to try it"  
[X Link](https://x.com/victormustar/status/1919772850514223388)  2025-05-06T15:14Z 23.2K followers, 221.2K engagements


"🤯 Its here: sub-10 second video generation is now real with LTX-Video-13B-distilled Try it now on Hugging Face"  
[X Link](https://x.com/victormustar/status/1922926343437561868)  2025-05-15T08:05Z 23.2K followers, 47K engagements


"Mac owners don't miss this: MLX LM is now integrated directly within Hugging Face 🤯 Run 4400+ LLMs locally on Apple Silicon at max speed no cloud no wait"  
[X Link](https://x.com/victormustar/status/1924510517311287508)  2025-05-19T17:00Z 23.2K followers, 208.5K engagements


"Chatterbox TTS is next level 🤯 seriously the speed at which open source AI is evolving right now is unreal. Demo available on Hugging Face"  
[X Link](https://x.com/victormustar/status/1928047822177292482)  2025-05-29T11:16Z 23.2K followers, 39.4K engagements


"Things are heating up 🔥🔥 Jan-nano by Menlo Research: a 4B Qwen3 finetune (using DAPO) for deep research tool use and search extraction On SimpleQA (agentic/MCP) Jan-nano scores [----] 🤯"  
[X Link](https://x.com/victormustar/status/1934354672086868457)  2025-06-15T20:57Z 23.2K followers, 25.2K engagements


"Nvidia cooked with PartPacker 3D Generation A new method to create 3D objects from a single image with each part separate and easy to edit 🔥 Demo available on Hugging Face"  
[X Link](https://x.com/victormustar/status/1934670798179992033)  2025-06-16T17:53Z 23.2K followers, 257.5K engagements


"Make sure to take some time to understand how hard NVIDIA cooked here 🤯🤯 Nvidia cooked with PartPacker 3D Generation A new method to create 3D objects from a single image with each part separate and easy to edit 🔥 Demo available on Hugging Face https://t.co/kY0RGmm6Op Nvidia cooked with PartPacker 3D Generation A new method to create 3D objects from a single image with each part separate and easy to edit 🔥 Demo available on Hugging Face https://t.co/kY0RGmm6Op"  
[X Link](https://x.com/victormustar/status/1934714416676605984)  2025-06-16T20:47Z 23.2K followers, 97.2K engagements


"Hugging Face Spaces the world's largest AI app directory is now MCP-compatible 🤯 Here turning an entire website into Ghibli in one shot with Claude Code to demonstrate what's now possible: instantly access any A I tool right in your LLM client"  
[X Link](https://x.com/victormustar/status/1935031748347236807)  2025-06-17T17:48Z 23.2K followers, 81.2K engagements


"DeepSite V2 just dropped (insane for Vibe Coding) 🤯 🔹Amazingly good at UI Design 🔹DeepSeek-R1 model now available 🔹Targeted Edits on any element 🔹Redesign any existing website 🔹100% open source (UI + models)"  
[X Link](https://x.com/victormustar/status/1935374796977766545)  2025-06-18T16:31Z 23.2K followers, 108.4K engagements


"DeepSeek-V3 is the 👑 of design Heres the result for luxury space rental. Does anyone have a point of comparison with other models/services"  
[X Link](https://x.com/victormustar/status/1935663566012453182)  2025-06-19T11:38Z 23.2K followers, 26.2K engagements


"what a time to be alive: a local 4B model (Jan-nano) + Web Search tool is already amazing to get the info you need with no overhead 🚀"  
[X Link](https://x.com/victormustar/status/1943656143143801203)  2025-07-11T12:58Z 23.2K followers, 61.3K engagements


"Ok I'm trying moonshotai/Kimi-K2-Instruct and wow. you werent lying 🤯 + being able to generate this in under [--] seconds is a game changer. 👇 Give it a go with this app (no account needed)"  
[X Link](https://x.com/victormustar/status/1945505576923889677)  2025-07-16T15:27Z 23.2K followers, 65.4K engagements


"New TTS bomb dropped on Hugging Face. It has multispeaker support and the output quality looks amazing 🤯"  
[X Link](https://x.com/victormustar/status/1948007505831100560)  2025-07-23T13:09Z 23.2K followers, 114.4K engagements


"this is crazy 👨🚀🚀"  
[X Link](https://x.com/victormustar/status/1950129944085864785)  2025-07-29T09:43Z 23.2K followers, 96.7K engagements


"OmniAvatar just dropped on Hugging Face and its really really good. Let the fun begin 😆 Sharing the Hugging Face demo below Omniavatar does offer quite intricate details when it comes to creating podcast style avatars with interesting facial expressions and hand gestures we also added a Zero GPU time estimation so if you can afford the GPU time the results are definitely worth it link 👇 https://t.co/VsfAseGSOH Omniavatar does offer quite intricate details when it comes to creating podcast style avatars with interesting facial expressions and hand gestures we also added a Zero GPU time"  
[X Link](https://x.com/victormustar/status/1952375379856863418)  2025-08-04T14:25Z 23.2K followers, 56.9K engagements


"Cool: Hugging Face is now supported in opencode select your favorite model from the Hub and it will consume your inference quota directly. I recommend starting with GLM-4.5 (great at coding + awesome at agentic tasks) Hugging Face PRO is the wildest $9/month deal in AI right now🤯 🔹 [--] min/day of H200 compute on Spaces ZeroGPU 🔹 1M free inference tokens from 15+ providers (Groq Cerebras etc.) 🔹 1TB private storage and more 🔹 More cool things. https://t.co/xGnZwLMEhH Hugging Face PRO is the wildest $9/month deal in AI right now🤯 🔹 [--] min/day of H200 compute on Spaces ZeroGPU 🔹 1M free"  
[X Link](https://x.com/victormustar/status/1952465745037639917)  2025-08-04T20:24Z 23.2K followers, 51.2K engagements


"Open Source AI is at its peak right now. just look at the current Hugging Face trending list: 🔹 Qwen/Qwen-Image-Edit 🔹 google/gemma-3-270m 🔹 tencent/Hunyuan-GameCraft-1.0 🔹 openai/gpt-oss-20b 🔹 zai-org/GLM-4.5V 🔹 deepseek-ai/DeepSeek-V3.1-Base 🔹 google/gemma-3-270m-it 🔹 Qwen/Qwen-Image 🔹 openai/gpt-oss-120b 🔹 janhq/Jan-v1-4B Models are already SOTA (or super close) now the only thing missing is probably the product layer 👀"  
[X Link](https://x.com/victormustar/status/1957867251060593058)  2025-08-19T18:08Z 23.2K followers, 62.1K engagements


"Cool: Nano Banana is now available on Hugging Face for all PRO users 👀"  
[X Link](https://x.com/victormustar/status/1960422110527926766)  2025-08-26T19:20Z 23.2K followers, 73.1K engagements


"🧑🍳 Microsoft cooked with VibeVoice-Large (MIT license) We are now super close to 100% open source NotebookLM 💫 (demo on Hugging Face)"  
[X Link](https://x.com/victormustar/status/1962540966948082035)  2025-09-01T15:40Z 23.2K followers, 30.1K engagements


"I'm sorry but this was peak AI"  
[X Link](https://x.com/victormustar/status/1962553279369326893)  2025-09-01T16:28Z 23.2K followers, 132.8K engagements


"Wan 2.2: First frame Last frame: Upload both as images to get excellent results. Amazing what open-source AI video can do now 😍 Demo available on Hugging Face"  
[X Link](https://x.com/victormustar/status/1963157112336392647)  2025-09-03T08:28Z 23.2K followers, 54.1K engagements


"Still a huge believer in small specialized models working together 🤗 Heres a good example: Tesslate/WEBGEN-4B-Preview: a 4B model (Qwen3-4B-Instruct-2507 finetune) specialized in Web UI generation. It's better at its task that most (huge) closed source models"  
[X Link](https://x.com/victormustar/status/1963315359756296270)  2025-09-03T18:57Z 23.2K followers, 29.8K engagements


"Interesting new 3D Gen demo on Hugging Face 👀 ReconViaGen: uses reconstruction priors and controlled diffusion to hallucinate missing views and deliver complete consistent 3D models from sparse overlap"  
[X Link](https://x.com/victormustar/status/1964999833493950955)  2025-09-08T10:30Z 23.2K followers, 51.9K engagements


"Tip: with your Hugging Face account you can try GLM-4.5 in your favorite coding tool for free to see if you like it (theres a good chance you will). It feels the coding agent frontier is now open-weights: GLM [---] costs only $3/month and is on par with Sonnet Kimi K2.1 Turbo is 3x speed 7x cheaper vs Opus [---] but as good Kimi K2.1 feels clean. The best model for me. GPT-5 is only good for complicated specs -- too slow. It feels the coding agent frontier is now open-weights: GLM [---] costs only $3/month and is on par with Sonnet Kimi K2.1 Turbo is 3x speed 7x cheaper vs Opus [---] but as good Kimi"  
[X Link](https://x.com/victormustar/status/1965046621105983567)  2025-09-08T13:36Z 23.2K followers, 81.9K engagements


"ChatGPT finally has MCP support 🔥"  
[X Link](https://x.com/victormustar/status/1965813896360632692)  2025-09-10T16:25Z 23.2K followers, 35.8K engagements


"another open source win: opencode + GLM [---] is basically Claude Code (used it all day) but insanely cheap + better TUI. And you can use it with your Hugging Face token now 🔥 Hugging Face PRO is the wildest $9/month deal in AI right now🤯 🔹 [--] min/day of H200 compute on Spaces ZeroGPU 🔹 1M free inference tokens from 15+ providers (Groq Cerebras etc.) 🔹 1TB private storage and more 🔹 More cool things. https://t.co/xGnZwLMEhH Hugging Face PRO is the wildest $9/month deal in AI right now🤯 🔹 [--] min/day of H200 compute on Spaces ZeroGPU 🔹 1M free inference tokens from 15+ providers (Groq"  
[X Link](https://x.com/victormustar/status/1973735580283625618)  2025-10-02T13:03Z 23.2K followers, 219.9K engagements


"Microsoft did something interesting here 👀 Unlike typical LLMs that are trained to play the role of the "assistant" in conversation we trained UserLM-8b to simulate the user role in conversation https://huggingface.co/microsoft/UserLM-8b https://huggingface.co/microsoft/UserLM-8b"  
[X Link](https://x.com/victormustar/status/1976345651656130614)  2025-10-09T17:54Z 23.2K followers, 193.7K engagements


"Introducing: HuggingChat Omni 💫 Select the best model for every prompt automatically 🚀 - Automatic model selection for your queries - [---] models available across [--] providers Available now all Hugging Face users. 100% open source"  
[X Link](https://x.com/victormustar/status/1978817795312808065)  2025-10-16T13:38Z 23.2K followers, 72.5K engagements


"Good to know: MinMax-M2 is one of the best model at UI design 👀 (its open source)"  
[X Link](https://x.com/victormustar/status/1983246160501346710)  2025-10-28T18:55Z 23.2K followers, 37.2K engagements


"New: Qwen-Image-2509-MultipleAngles. Very solid model and probably a lot of creative use cases to find with it. Free demo available on Hugging Face"  
[X Link](https://x.com/victormustar/status/1985657661963075943)  2025-11-04T10:37Z 23.2K followers, 749.9K engagements


"State-of-the-art New: Qwen-Image-2509-MultipleAngles. Very solid model and probably a lot of creative use cases to find with it. Free demo available on Hugging Face https://t.co/uYdIoBtXsP New: Qwen-Image-2509-MultipleAngles. Very solid model and probably a lot of creative use cases to find with it. Free demo available on Hugging Face https://t.co/uYdIoBtXsP"  
[X Link](https://x.com/victormustar/status/1985658210057925096)  2025-11-04T10:39Z 23.2K followers, 662.1K engagements


"New Llama.cpp UI is a blessing for the local AI world 🌎 - Blazing fast beautiful and private (ofc) - Use 150000+ GGUF models in a super slick UI - Drop in PDFs images or text documents - Branch and edit conversations anytime - Parallel chats and image processing - Math and code rendering - Constrained generation with JSON schema supported Easy setup + open-source + community-built 🔥 A detailed look into the new WebUI of llama.cpp https://t.co/o26pF28pM7 A detailed look into the new WebUI of llama.cpp https://t.co/o26pF28pM7"  
[X Link](https://x.com/victormustar/status/1985742628776706151)  2025-11-04T16:15Z 23.2K followers, 163.9K engagements


"Maya Research drops a TTS bomb: Maya1 🔥 🔹Create any voice: from a 20s British girl to a full-blown demon 🔹Great at emotions: laugh cry whisper rage sigh gasp & more 🔹Extremely fast: 3B parameters runs on a single GPU 🔹Outperforms top proprietary models Demo available on Hugging Face"  
[X Link](https://x.com/victormustar/status/1987828962047738187)  2025-11-10T10:25Z 23.2K followers, 52.9K engagements


"@flowVSgravity Yes the llamacpp webui is great"  
[X Link](https://x.com/victormustar/status/1999571974863847931)  2025-12-12T20:08Z 23.2K followers, [---] engagements


"Who is running this locally (309B MOE with A15B 🔥) I need to know this could be it 👀 https://huggingface.co/XiaomiMiMo/MiMo-V2-Flash https://huggingface.co/XiaomiMiMo/MiMo-V2-Flash"  
[X Link](https://x.com/victormustar/status/2001030892362338413)  2025-12-16T20:45Z 23.2K followers, 11K engagements


"Microsoft TRELLIS.2 is here 🔥 Single image textured 3D mesh 4B params flow-matching transformer Up to [----] resolution Open weights MIT licensed Demo available on Hugging Face"  
[X Link](https://x.com/victormustar/status/2001207429237703036)  2025-12-17T08:26Z 23.2K followers, 199.1K engagements


"🚀 Supertonic [--] TTS is already here - Runs entirely on-device - Generates [--] sec of audio in [-----] sec - Super impressive quality - 66M params no cloud no API calls - [--] languages: EN KO ES PT FR Demo available on Hugging Face"  
[X Link](https://x.com/victormustar/status/2010662954086408494)  2026-01-12T10:39Z 23.2K followers, 36.5K engagements


"@TommyFalkowski llamacpp up to date + new checkpoints GLM-4.7-Flash GGUFs now produce significantly better outputs after recent llama.cpp bug fixes. We reconverted and updated the GGUFs. Run 4-bit locally on 18GB RAM. To get fixes re-download & use inference parameters by @Zai_org. Updated GGUFs: https://t.co/aTuUxu32z3 https://t.co/K74kMzlbIK GLM-4.7-Flash GGUFs now produce significantly better outputs after recent llama.cpp bug fixes. We reconverted and updated the GGUFs. Run 4-bit locally on 18GB RAM. To get fixes re-download & use inference parameters by @Zai_org. Updated GGUFs:"  
[X Link](https://x.com/victormustar/status/2013982145413353650)  2026-01-21T14:28Z 23.2K followers, [--] engagements


"@HaareBlond i dont think lm-studio has latest llamacpp fix (probably only latest release) - check if it's this setting: https://x.com/UnslothAI/status/2013966866646180345 GLM-4.7-Flash GGUFs now produce significantly better outputs after recent llama.cpp bug fixes. We reconverted and updated the GGUFs. Run 4-bit locally on 18GB RAM. To get fixes re-download & use inference parameters by @Zai_org. Updated GGUFs: https://t.co/aTuUxu32z3 https://t.co/K74kMzlbIK https://x.com/UnslothAI/status/2013966866646180345 GLM-4.7-Flash GGUFs now produce significantly better outputs after recent llama.cpp"  
[X Link](https://x.com/victormustar/status/2014361179246776411)  2026-01-22T15:35Z 23.2K followers, [--] engagements


"@cocktailpeanut @openclaw it's not possible to run it directly on the R1 (didn't follow this)"  
[X Link](https://x.com/victormustar/status/2017294461261299794)  2026-01-30T17:50Z 23.2K followers, [----] engagements


"@antirez will definitely try personally I've been using it with a custom command from claude only for reviews and it's working quite well"  
[X Link](https://x.com/victormustar/status/2017344549807366373)  2026-01-30T21:09Z 23.2K followers, [----] engagements


"DeepSeek is silently building all the blocks they need to ship something crazy for R2 instead of rushing it 🐳🤩 The whale is BACK 👀👀👀 https://t.co/BsGKPw4r7r The whale is BACK 👀👀👀 https://t.co/BsGKPw4r7r"  
[X Link](https://x.com/victormustar/status/1994000378371989716)  2025-11-27T11:08Z 23.2K followers, 33.9K engagements


"@thinkfulearn @UnslothAI @huggingface yes it's quite simple install llamacpp then llama-server comes with a nice chat UI https://huggingface.co/unsloth/MiniMax-M2.1-GGUFlocal-app=llama.cpp https://huggingface.co/unsloth/MiniMax-M2.1-GGUFlocal-app=llama.cpp"  
[X Link](https://x.com/victormustar/status/2005040686979318094)  2025-12-27T22:18Z 23.2K followers, [----] engagements

Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing