/technology — bitchan

A general for vibe coding, agentic engineering, coding agents, and shipping code with LLMs. You use Git — right, anon? ## What “vibe coding” is, and how to do it https://simonwillison.net/2025/Mar/19/vibe-coding/ https://simonwillison.net/2025/Mar/11/using-llms-for-code/ ## Related generals >>>/g/lmg/ ---- ## Frontier models using fully-general tooling — start here if you have $20 or so https://claude.com/product/claude-code — generally better currently https://developers.openai.com/codex/cli h…⬆↩ 30

Having Fable as a Socratic partner in fielding ideas and exploring counterpoints is so great. Fable 5.5 will BTFO the field if it's only ~2X as expensive in practice as 5.1+

⬆ upvote 0 ↩ reply 0

Troons have used open source to fund their hrt and drug addictions for years. The donators develop a parasocial relationship with the troon. Vibecoders take this away.+

⬆ upvote 0 ↩ reply 0

>run through weekly limit in 2 days >get second account >2 hours in >already at 10% weekly limit hope it lasts until monday+

⬆ upvote 0 ↩ reply 0

+ 27

/lmg/ - a general dedicated to the discussion and development of local language models. News >(10/27) NVidia releases Nemotron 3.5 Transcribe 30B-A4B https://hf.co/nvidia/NVIDIA-Nemotron-3.5-Transcribe-30B-A4B >(10/26) XHToken releases Spark-X2.6-1.7B: https://hf.co/XHToken/Spark-X2.6-1.7B >(10/26) LiquidAI releases LFM3 1.7B and 2B: https://hf.co/collections/LiquidAI/lfm3 >(10/26) Baidu Unlimited-OCR-2 3B released https://hf.co/baidu/Unlimited-OCR-2 >(10/23) NVIDIA Nemotron 3.5 Bolt E4B, postr…⬆↩ 32

>Imagine dumping $20K+ into a shitty locked+soldered computer for the privilege to run cope quants at barely usable speeds. Wouldn't be me.+

⬆ upvote 0 ↩ reply 0

Just invest in a data center and have some compute diverted for personal use.+

⬆ upvote 0 ↩ reply 0

How many tok / s on deepseek flash can I expect with the 512GB version mac running with the 36 cores option?+ 1

⬆ upvote 0 ↩ reply 1

30-40tok/s if you run a cope quant.+

⬆ upvote 0 ↩ reply 0

+ 29

/lmg/ - a general dedicated to the discussion and development of local language models. News >(10/23) NVIDIA Nemotron 3.5 Bolt E4B, postrained model: https://hf.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-E4B-BF16 >(10/23) MiniMax-M3.1-Flash 428B-A23B released: https://hf.co/MiniMaxAI/MiniMax-M3.1-Flash >(10/21) Recursive Flint 40B-A2.5B released: https://recursive.com/updates/introducing-recursive-flint >(10/21) Korean Raon-1 mini 30B-A4B released: https://hf.co/KRAFTON/Raon-1-mini-preview >(10/20) Orn…⬆↩ 29

Recent Highlights from the --Anons skeptical of Cohere's report alleging Qwen data exfiltration: >110169543 >110169564 >110169580 >110169621 >110169740 >110169641 >110169660 >110169678 >110169671 >110169778 >110169816 >110169873 --GeForce FP16/BF16 unlock boosts prefill, token gen mostly unaffected: >110166270 >110166349 >110166389 >110166441 --Reducing Nvidia driver reserved VRAM on Linux and Windows: >110166849 >110166913 >110166930 >110166969 >110167030 >110167049 >110167175 >110167181 >1101…+

⬆ upvote 0 ↩ reply 0

> $832B in non-cancelable obligations dariosissy...+ 1

⬆ upvote 0 ↩ reply 1

>inb4 too big to fail+

⬆ upvote 0 ↩ reply 0

https://desuarchive.org/g/thread/109690289/#q109693327 >all your data are belong to us+

⬆ upvote 0 ↩ reply 0

+ 26

A general for vibe coding, agentic engineering, coding agents, and shipping code with LLMs. You use Git — right, anon? ## What “vibe coding” is, and how to do it https://simonwillison.net/2025/Mar/19/vibe-coding/ https://simonwillison.net/2025/Mar/11/using-llms-for-code/ ## Related generals >>>/g/lmg/ ---- ## Frontier models using fully-general tooling — start here if you have $20 or so https://claude.com/product/claude-code — generally better currently https://developers.openai.com/codex/cli h…⬆↩ 35

that's so hot+

⬆ upvote 0 ↩ reply 0

Dad works at Anthropic, he said that Fable 5.5 releases today with a reset+

⬆ upvote 0 ↩ reply 0

sex with sam topman+

⬆ upvote 0 ↩ reply 0

+ 32

what does the kroma lora do for krea2 and why should i not take it+

⬆ upvote 0 ↩ reply 0

Also, turbo lora rapes details.+

⬆ upvote 0 ↩ reply 0

erm excuse me, old thread isn't at bump limit yet+

⬆ upvote 0 ↩ reply 0

+ 18

/lmg/ - a general dedicated to the discussion and development of local language models News >(10/23) NVIDIA Nemotron 3.5 Bolt E4B, postrained model: https://hf.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-E4B-BF16 >(10/23) MiniMax-M3.1-Flash 428B-A23B released: https://hf.co/MiniMaxAI/MiniMax-M3.1-Flash >(10/21) Recursive Flint 40B-A2.5B released: https://recursive.com/updates/introducing-recursive-flint >(10/21) Korean Raon-1 mini 30B-A4B released: https://hf.co/KRAFTON/Raon-1-mini-preview >(10/20) Orni…⬆↩ 32

stick my finger in her nostril+

⬆ upvote 0 ↩ reply 0

you're giving me too many things lately you're all i need oh oh+

⬆ upvote 0 ↩ reply 0

Recent Highlights from the --Comparing Chinese models and hardware requirements for running DeepSeek: >110156576 >110156582 >110156589 >110156599 >110156623 >110156619 >110156671 >110156994 >110157020 >110157034 >110157044 >110156592 >110156605 >110156953 >110156977 >110157227 >110157287 >110157311 >110157329 >110156683 --Comparing DSV4.1 and qwenext Q4 for Godot 4 game development: >110154561 >110154569 >110154591 >110154626 >110154671 >110154739 >110154779 >110155637 >110154915 >110154968 >11…+

⬆ upvote 0 ↩ reply 0

+ 29

Motorola Signature 27 Edition! >What phone has X and Y feature? https://www.gsmarena.com/search.php3 https://www.kimovil.com/en/compare-smartphones https://phonedb.net/index.php?m=device&s=query >Reviews https://www.gsmarena.com https://www.phonearena.com https://www.notebookcheck.net >Recommended phones https://www.gsmarena.com/smartphone_buyers_guide-review-2036.php >Frequency Checkers https://www.frequencycheck.com https://kimovil.com/en/frequency-checker https://willmyphonework.net >Visual…⬆↩ 31

>>110149755 Based OP >>110149833 I might buy one too and I'm the anon with all the new flagships. :^)+ 1

⬆ upvote 0 ↩ reply 1

>>110149851 How many flahship phones do you own?+

⬆ upvote 0 ↩ reply 0

>>110149755 I don't like the camera layout at all but the hardware and grapheneOS supports cool.+

⬆ upvote 0 ↩ reply 0

I told someone I wanted to get this, and she’s all like “nice calculator, dweeb” why are foids so dumb and evil?+

⬆ upvote 0 ↩ reply 0

+ 28

A general for vibe coding, agentic engineering, coding agents, and shipping code with LLMs. You use Git — right, anon? ## What “vibe coding” is, and how to do it https://simonwillison.net/2025/Mar/19/vibe-coding/ https://simonwillison.net/2025/Mar/11/using-llms-for-code/ ## Related generals >>>/g/lmg/ ---- ## Frontier models using fully-general tooling — start here if you have $20 or so https://claude.com/product/claude-code — generally better currently https://developers.openai.com/codex/cli #…⬆↩ 28

Google LOST!!!+

⬆ upvote 0 ↩ reply 0

https://x.com/claudeai/status/2115467192276025500 The endgame is to make these kind of shitty tools a SaaS after going public The golden age of frontier models are going to end...+

⬆ upvote 0 ↩ reply 0

you'll never be a real programmer+

⬆ upvote 0 ↩ reply 0

+ 25

/lmg/ - a general dedicated to the discussion and development of local language models. News >(10/29) Mistral Large 4 (Le Chonk) 1T-A52B released: https://mistral.ai/news/mistral-large-4-oss >(10/27) NVidia releases Nemotron 3.5 Transcribe 30B-A4B: https://hf.co/nvidia/NVIDIA-Nemotron-3.5-Transcribe-30B-A4B >(10/26) XHToken releases Spark-X2.6-1.7B: https://hf.co/XHToken/Spark-X2.6-1.7B >(10/26) LiquidAI releases LFM3 1.7B and 2B: https://hf.co/collections/LiquidAI/lfm3 >(10/26) Baidu Unlimited…⬆↩ 28

omg it rin+

⬆ upvote 0 ↩ reply 0

Recent Highlights from the --Mistral Large 4 open weight release and licensing discussion: >110193924 >110193945 >110194061 >110194069 >110195025 --Methods for sandboxing LLM agents to prevent system damage: >110191862 >110191875 >110191893 >110191901 >110191902 >110191923 >110191939 >110191935 >110191943 >110191973 >110191959 >110191996 >110192210 >110192243 >110193143 >110193370 --Research on emotional instability and frustration in Gemma models: >110193276 >110193301 >110193328 >110193459 >1…+ 1

⬆ upvote 0 ↩ reply 1

>Research on emotional instability and frustration in Gemma models: Best part of the threat. Somebody should reproduce it with Gemma 4.+

⬆ upvote 0 ↩ reply 0

how long before Opus 5.5 tier coder will be available for local LLM ? Or this level is rather unattainable on normal consumer Pcs?+

⬆ upvote 0 ↩ reply 0

+ 25

/lmg/ - a general dedicated to the discussion and development of local language models. News >(10/29) Mistral Large 4 (Le Chonk) 1T-A52B released: https://mistral.ai/news/mistral-large-4-oss >(10/27) NVidia releases Nemotron 3.5 Transcribe 30B-A4B: https://hf.co/nvidia/NVIDIA-Nemotron-3.5-Transcribe-30B-A4B >(10/26) XHToken releases Spark-X2.6-1.7B: https://hf.co/XHToken/Spark-X2.6-1.7B >(10/26) LiquidAI releases LFM3 1.7B and 2B: https://hf.co/collections/LiquidAI/lfm3 >(10/26) Baidu Unlimited…⬆↩ 30

how can i nyomaxx with gemmy? what system prompts should i use?+

⬆ upvote 0 ↩ reply 0

> Yeah the J++ lens barely beats the logit lens in most layers. Anyway check out the latest update from this morning on github, my 5090 spent all last night doing the first full layer neuron census. If any anons with fat GPU wants to help me finish the other layers, that would be awesome, otherwise it will take me months without a runpod instance. https://github.com/whp199/logos-theodolite+ 1

⬆ upvote 0 ↩ reply 1

Let's start an /lmg/ AI lab and ask for funding. I'll make the logo.+

⬆ upvote 0 ↩ reply 0

So since Claude got cucked again after finding an actual use of AI.... .... Can Local Models port PC games ????+ 1

⬆ upvote 0 ↩ reply 1

Yes+

⬆ upvote 0 ↩ reply 0

+ 27

Users of all levels are welcome to ask questions about GNU/Linux and share their experiences. *** Please be civil — notice the "Friendly" in every Friendly GNU/Linux Thread. *** Before asking for help, please check our list of resources. >Which distro should I choose? https://nosystemd.org https://systemdfree.com >What are some cool programs? https://suckless.org https://sl.netcat.uk >What are some cool terminal commands? https://cheat.sh >Where can I learn the command line? RTFM >Where can I learn more about Free Software? https://fsf.org https://gnu.org https://stallman.org >How to break out of the botnet? https://kycnot.me https://privacyguides.org GNU/Linux Games: >>>/vg/lgg⬆↩ 30

Fedora KDE: Is there a way to summon a temporary terminal that kills itself when it's no longer focused? I just went on an adventure with the bot and it spit out some bash script that will probably kill the wrong process someday. Usecase: temporary ncmpcpp instance I summon with a hotkey+

+ 27

A general for vibe coding, agentic engineering, coding agents, and shipping code with LLMs. You use Git — right, anon? ## What “vibe coding” is, and how to do it https://simonwillison.net/2025/Mar/19/vibe-coding/ https://simonwillison.net/2025/Mar/11/using-llms-for-code/ ## Related generals >>>/g/lmg/ ---- ## Frontier models using fully-general tooling — start here if you have $20 or so https://claude.com/product/claude-code — generally better currently https://developers.openai.com/codex/cli h…⬆↩ 20

sex with sam altman rape snailcats kill dario+

⬆ upvote 0 ↩ reply 0

I vibecoded my own Wardogs cheat+

⬆ upvote 0 ↩ reply 0

new zogwatch dropped https://youtu.be/OP5GDiLZ1Nw+

⬆ upvote 0 ↩ reply 0

+ 17

No collage edition (don't know how to do it, sowwy) >UIs ComfyUI: https://github.com/comfyanonymous/ComfyUI SwarmUI: https://github.com/mcmonkeyprojects/SwarmUI KritaAI: https://github.com/Acly/krita-ai-diffusion Forge Classic/Neo: https://github.com/Haoming02/sd-webui-forge-classic InvokeAI: https://github.com/invoke-ai/InvokeAI >Image Models IL: https://rentry.org/illustrious_checkpoints Anima: https://huggingface.co/circlestone-labs/Anima Krea: https://huggingface.co/krea/Krea-2-Turbo NovelA…⬆↩ 30

I too do not know how the collage, unfortunately+ 1

⬆ upvote 0 ↩ reply 1

https://rentry.org/neo_collage+

⬆ upvote 0 ↩ reply 0

Kek I'm getting some interspecies reviewers vibes here+

⬆ upvote 0 ↩ reply 0

what's the point of posting in this fucking general if OP is not gonna take the time to make the collage+ 1

⬆ upvote 0 ↩ reply 1

Are you asking that as a philosophical? I liked your gens from last thread+

⬆ upvote 0 ↩ reply 0

+ 27

Science edition How to request advice: >Budget >Intended use (media, source, environment) >Frequency response preference and music examples >Past gear and your thoughts on them FAQ: >Where do I buy IEMs? Amazon, Aliexpress, Linsoul, Hifigo, Shenzhenaudio >Shopping Guide (IEMs, PMPs, Cables, Ear Tips, etc.): https://consoomer-guide.pages.dev/ >/iemg/'s Journal: https://iemgazette.pages.dev/ >EQ Guide (EQ 101, Targets, Myths & Misconceptions, Papers, etc.): https://iem-eq-guide.pages.dev/ >Measur…⬆↩ 24

/iemg/ Dogma Check Tired of the hivemind? Here’s your prescribed sanity dose: >Crinacle is not your priest. Rankings are a guide, not divine scripture. Your ears > his spreadsheet. >Harman and JM-1 ≠ perfect. It’s a preference curve, not a law of physics. Some genres need deviation. >Chi-Fi isn’t magic. Great value, yes. Immune to QC issues and hype cycles? Nope. >Graphs are not sound. They don’t show staging, dynamics, or fit. Use your ears, not just squigs. >BA ≠ metallic, DD ≠ perfect. Imple…+ 1

⬆ upvote 0 ↩ reply 1

Doesn't know what EQ is btw.+

⬆ upvote 0 ↩ reply 0

is there any point using the entire setup for IEMs? Looks like you only need the ears+ 1

⬆ upvote 0 ↩ reply 1

For IEMs you don't even need the ears. You officially have no ballz from now on.+

⬆ upvote 0 ↩ reply 0

You fags lied to me. I can't hear any difference with or without a DAC (plugging it directly to my PC).+ 1

⬆ upvote 0 ↩ reply 1

If it's a 1DD IEM and there's no hiss from the PC then there won't be a difference. Some multidrivers like Nova don't change much from output impedance but most of them do.+

⬆ upvote 0 ↩ reply 0

+ 21

A general for vibe coding, agentic engineering, coding agents, and shipping code with LLMs. You use Git — right, anon? ## What “vibe coding” is, and how to do it https://simonwillison.net/2025/Mar/19/vibe-coding/ https://simonwillison.net/2025/Mar/11/using-llms-for-code/ ## Related generals >>>/g/lmg/ ---- ## Frontier models using fully-general tooling — start here if you have $20 or so https://claude.com/product/claude-code — generally better currently https://developers.openai.com/codex/cli #…⬆↩ 18

I just got a whole month of chatgpt plus for free(tm). How are codex's astra limits with this plan? I want to build a few things but mostly a blender MCP pipeline and game mods. Can I do that with this sub?+ 1

⬆ upvote 0 ↩ reply 1

don't use astra with plus, it burns usage too quick stick to 6.1 sol medium, limits are pretty generous there gpt modelling is mid compared to opus you'll probably try astra anyway though, but i guess you gotta experience your 5hr limit evaporating in 20 minutes for your self+

⬆ upvote 0 ↩ reply 0

Does anybody have good skills (I know there's a link in the OP), plugins, hooks, custom configuration, or MCPs for their vibe coding? I'm just curious about favorites for programming and general code generation.+ 1

⬆ upvote 0 ↩ reply 1

The skills you make yourself based on your own workflows and failures you observe+

⬆ upvote 0 ↩ reply 0

Later, I'm doing a house retrofit rn+ 1

⬆ upvote 0 ↩ reply 1

Using what?+

⬆ upvote 0 ↩ reply 0

+ 15

Discussion and Development of Local Image, Video, and Audio Models https://rentry.org/ldg-lazy-getting-started-guide >UI ComfyUI: https://github.com/comfyanonymous/ComfyUI SwarmUI: https://github.com/mcmonkeyprojects/SwarmUI SDWebUI: https://rentry.org/ldg-lazy-getting-started-guide#the-stable-diffusion-web-ui-lineage Wan2GP: https://github.com/deepbeepmeep/Wan2GP Neural-Pixel (sd.cpp): https://github.com/Luiz-Alcantara/Neural-Pixel >Checkpoints, LoRAs, & Upscalers https://huggingface.co/models…⬆↩ 21

>debo is spiraling again Poor disabo mad that he can't use the singular anon cope that has been proven false for years now.+

⬆ upvote 0 ↩ reply 0

>normalfag retards cheering on the ai bubble popping as if ai will disappear baga+

⬆ upvote 0 ↩ reply 0

TY 4 BAKE+

⬆ upvote 0 ↩ reply 0

+ 18