Marcel Joist

Managing Director & 11x Certified Salesforce Application Architect and Mulesoft Developer at H+W Consult GmbH LinkedIn :

[Marcel Joist @MJoist]

@QwenDevs @MiaAI_lab Do it

[Marcel Joist @MJoist]

@Alibaba_Qwen @NVIDIARTXSpark Just, thank you !

[Marcel Joist @MJoist]

@caseyjp11 @Alibaba_Qwen

[Marcel Joist @MJoist]

@MiaAI_lab @ivanfioravanti @NVIDIAAI Thank you

[Marcel Joist @MJoist]

@EvenRealities I really like it so much.

[Marcel Joist @MJoist]

@EvenRealities Please sent it i am waiting Ordered 31.01

[Justin Kalland @justinkalland]

LLM speed on my MacBook M5 Max (128GB):

• Qwen3.5-35B-A3B (Q6): 74 tok/s
• Nemotron-3 Super (Q4): 24 tok/s
• Qwen3-Coder-Next (6-bit): 67 tok/s
• Llama 3.3 8B Instruct (Q4): 99 tok/s

On my M1, Llama 3.3 was my go-to for most local tasks at about 20 tok/s. On the M5 Max, it's hitting 99 tok/s.

Qwen3.5 feels like a huge upgrade and is my favorite so far.

Qwen3-Coder-Next is surprisingly good at dev tasks, although I'll probably stick with GPT-5.4 for most.

I'm also impressed by Nemotron-3 Super, but its personality feels a bit too dry.

[Unsloth AI @UnslothAI]

Qwen3.5 is now updated with improved tool-calling & coding performance!

Run Qwen3.5-35B-A3B on 22GB RAM.
See improvements via Claude Code, Codex.

We also benchmarked GGUFs & removed MXFP4 layers from 3 quants.

GGUFs: https://t.co/4lSce5zZbO
Analysis: https://t.co/rHZK8JWdYM

[Claude @claudeai]

New in Cowork: scheduled tasks.

Claude can now complete recurring tasks at specific times automatically: a morning brief, weekly spreadsheet updates, Friday team presentations.

[Claude @claudeai]

Admins can create private plugin marketplaces to distribute them across the org.

A unified "Customize" menu also gives you more control over plugins, skills, and connectors in one place.

[Hyper3D by Deemos @DeemosTech]

🤔First Text/Image-to-3D Skill?

🔥Claude Code just made a selfie for @grok & itself instantly with Rodin #3D #Skill 😂

🚀Quick Setup:
/plugin marketplace add DeemosTech/rodin3d-skills
/plugin install rodin3d-skill@rodin3d-skills

Free to use! Just ask #Claude to generate🎨

[Hyper3D by Deemos @DeemosTech]

GPT-5.3-Codex + Image-to-3D Skill
Just generated a cute avatar for @sama

[Marcel Joist @MJoist]

@"Das eingefrorene Gehirn: Warum wir von AGI weiter entfernt sind als viele denken!"https://t.co/PX4bSxeGRS auf @LinkedIn ⁦@salesforce⁩ ⁦@OpenAI⁩

[Min Choi @minchoi]

Anthropic literally dropped the Prompt Guide to 10x your Claude Opus/Sonnet 4.5

10 best practices you should know. Bookmark this. 👇

[Kilo (acq. by Anaconda) @kilocode]

We compared GPT-5.1 vs Gemini 3.0 vs Claude Opus 4.5 across 3 coding tasks.

The results reveal pros and cons for each model: https://t.co/gneNia4GBL

[Marcel Joist @MJoist]

@SadlyItsBradley One question is screensharing in meetings possible to use like Google meet or slack and ms teams app? Can you share a view and a single window ?

[🚨 AI News | TestingCatalog @testingcatalog]

BREAKING 🚨: Claude Skills are rolling out to all paid users! Skills are packaged and shareable instructions to guide Claude on how to approach your workflows.

[cat @_catwu]

Haiku 4.5 is a workhorse that makes the coding experience in Claude Code feel really fast.

While Sonnet 4.5 remains the default, Haiku 4.5 now powers the Explore subagent which can rapidly gather context on your codebase to build apps even faster.

[Marcel Joist @MJoist]

https://t.co/78TByy7cld

[Elon Musk @elonmusk]

Upload any photos from your phone or take a picture of an old photo and turn it into a video in 20 seconds with the Grok app!