David Garcia

[David Garcia @iamdagy]

@sunbaolong_2001 Would it be available in the comfyui cloud nodes ?

[David Garcia @iamdagy]

@PurzBeats @davidpantera_ When would it be available in comfyui cloud ? 😬😅 looks fun

[David Garcia @iamdagy]

@NousResearch @Teknium Hi, what is the recommmended Max context length for hermes ? 64k 100k 200k ? Si tool calls and chat and compaction and everything works as ideally expected

[David Garcia @iamdagy]

@ComfyUI the link seems broken. where can we see the replay ?

[David Garcia @iamdagy]

@outsource_ @LottoLabs That would be awesome

[David Garcia @iamdagy]

@marcsh @LottoLabs How are searxhibg trough the model ? Court you give more details . Thanks

[David Garcia @iamdagy]

@NVIDIAGeForce PRAGMATA RTX

[David Garcia @iamdagy]

@WuMinghao_nlp Looks great. Also looking really forward for the dense models

[David Garcia @iamdagy]

@TheAhmadOsman @ClementDelangue Is it multimodal like qwen 3.5?

[David Garcia @iamdagy]

@bnjmn_marie Is kaitchup W4A16 fast as NVFP4 on an RTX5090 Blackwell ?

[David Garcia @iamdagy]

@MariaFdaCabal Que los publiquen como en usa publicaron los de epstein !

[David Garcia @iamdagy]

@Ali_TongyiLab What is the default/recommended image size : does it do 2k, 4k ?

[David Garcia @iamdagy]

@wildmindai How does Z image compares to qwen image ? I have used qwen image but haven’t used Z image ..

[David Garcia @iamdagy]

@replicate Great

[David Garcia @iamdagy]

@Alibaba_Qwen Great news thanks ! 🥳.. wondering how does the 8B compares to the 30B version ? Thanks for making this great models open to all

[David Garcia @iamdagy]

@cherry_cc12 Qwen 3 VL 🔥

[David Garcia @iamdagy]

@NVIDIAGeForce GeoForce RTX 5090

[David Garcia @iamdagy]

@ComfyUI Hi, How did you set it up to run in a rtx5090 . Im trying but I’m getting out of memory . I have to either disable the 4 steps Lora’s or offload the clip to the cpu

[David Garcia @iamdagy]

@thdxr Wondering what is the best coding model to run offline with opencode on a 18gb ram Mac book pro m3 pro ? Had a long fly and try a bunch of 7b parameters with lmstudio and ollama but none use tools right, kind of imposible to use 😅 I want to be better prepared for next flight