Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

A 397B parameter model is running on somebody's desk right now. Not a datacenter. A desk. Qwen shipped their new 397B flagship and the usual take was instant: nobody runs this at home. One creator took that personally. He chained 4 DGX Sparks into a single cluster. 128GB of... memory each, linked with the kind of networking you normally see in server rooms. The clip is 76 seconds. Watch the memory readouts around the halfway point. All boxes pinned at 112 of 119GB. The model barely fits, and it fits. Then he opens a terminal on his Mac and benchmarks the whole thing like it's a cloud API. Same commands, same endpoint, except the endpoint is 3 feet away. I paused twice on the monitoring screen just to count the memory bars. A year ago this class of model lived behind an API key and a waitlist. The datacenter lost its monopoly. It just doesn't know yet.show more

Vettan

35,041 subscribers

15,542 views • 3 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

16 Comments

Misato3 months ago

barely fits is my favorite kind of homelab build

Vettan3 months ago

Yeah my too

Dima T.3 months ago

F#ck !)) This breaks my brain. A year ago: "this model only lives in the cloud." Now: guy ran it on his desk. Works like a server. Commands from Mac like it's the cloud. But it's just his desk.

Mateo Delgado3 months ago

4 sparks is 512GB total right? that's like $8-10k used. still not "desk" money for most people but yeah the impossible-at-home takes age in real time now

Shelby Cobb3 months ago

what's the exact name of hte camera/lens used for this? the depth of field is flawless

eryenrye3 months ago

AI infrastructure is becoming more accessible every year. What used to require a datacenter is slowly becoming possible for smaller teams and individual builders

sofi t,3 months ago

mac mini with 24gb ram fighting for its life

Matt3 months ago

😍😎

mcguire kelly3 months ago

the emotional range this video just took me through

GryptoNews3 months ago

barely fits is my favorite kind of

Clint Brasby3 months ago

the subtle detail that nobody noticed changes everything

Radar 24h3 months ago

Innovation knows no limits. Memory bars tell the story.

Maxl0nG3 months ago

so all attacks mac and mini will get nvidia video cards?

Arie Mctiernan3 months ago

the story behind this is honestly way more interesting than the video

Tasha Inkavesvanitc3 months ago

can someone explain the physics of what just happened at the end

Shubh Thorat3 months ago

underrated part of this: quantization and unified memory quietly did more for local AI than any single model release this year

Related Videos

TWO BOXES THE SIZE OF A MAC MINI JUST RAN A 235 BILLION PARAMETER MODEL ON A DESK It is two NVIDIA DGX Spark units linked by a single cable. A year ago a model this size meant renting a GPU cluster by the hour. Now it sits next to your monitor for around $8,000. Here is the twist most people miss. Linking them does not create one shared 256GB memory pool. The model is split across both boxes, and that is the only reason a 235B model fits at all. It answers at roughly 10 tokens per second, and both chips sit at just 74 degrees while sipping around 50 watts. Every token stays on the desk. Nothing touches a cloud, and nothing leaves the room. The ceiling for what you can run at home just jumped from 70B to 235B. Bookmark this & Watch it run ↓

TWO BOXES THE SIZE OF A MAC MINI JUST RAN A 235 BILLION PARAMETER MODEL ON A DESK It is two NVIDIA DGX Spark units linked by a single cable. A year ago a model this size meant renting a GPU cluster by the hour. Now it sits next to your monitor for around $8,000. Here is the twist most people miss. Linking them does not create one shared 256GB memory pool. The model is split across both boxes, and that is the only reason a 235B model fits at all. It answers at roughly 10 tokens per second, and both chips sit at just 74 degrees while sipping around 50 watts. Every token stays on the desk. Nothing touches a cloud, and nothing leaves the room. The ceiling for what you can run at home just jumped from 70B to 235B. Bookmark this & Watch it run ↓

slash1s

101,317 views • 4 months ago

$FOUR BOXES ON A DESK ARE NOW A FRONTIER AI CLUSTER This is four NVIDIA DGX Spark units wired together, sitting on a single desk and running Qwen3-235B end to end. Each box holds 128GB of unified memory. Stacked together, that is half a terabyte of memory pooled across the cluster, all from hardware you could fit in a backpack. A year ago a setup like this lived in a data center behind a badge reader. Now it runs next to a monitor and a coffee cup. Here is the part people get wrong. The four boxes do not magically merge into one giant GPU. The model is sharded across all of them, and they pass data over a fast link while every box does its slice of the work. That is the whole trick. No single box could hold a 235B model, but split four ways, it runs. The machines stay quiet and barely warm while doing it, sipping a fraction of the power a rack of server GPUs would burn. The wall between you and a frontier-class model used to be a cloud bill. Now it is just desk space. Bookmark this & Watch it run ↓$

FOUR BOXES ON A DESK ARE NOW A FRONTIER AI CLUSTER This is four NVIDIA DGX Spark units wired together, sitting on a single desk and running Qwen3-235B end to end. Each box holds 128GB of unified memory. Stacked together, that is half a terabyte of memory pooled across the cluster, all from hardware you could fit in a backpack. A year ago a setup like this lived in a data center behind a badge reader. Now it runs next to a monitor and a coffee cup. Here is the part people get wrong. The four boxes do not magically merge into one giant GPU. The model is sharded across all of them, and they pass data over a fast link while every box does its slice of the work. That is the whole trick. No single box could hold a 235B model, but split four ways, it runs. The machines stay quiet and barely warm while doing it, sipping a fraction of the power a rack of server GPUs would burn. The wall between you and a frontier-class model used to be a cloud bill. Now it is just desk space. Bookmark this & Watch it run ↓

slash1s

19,473 views • 4 months ago

THIS AI DEVELOPER RUNS 120B PARAMETER MODELS LOCALLY AND HIS MACBOOK PRO CAN’T EVEN OPEN THEM he bought a DGX Spark, gave it the model, went back to his desk by end of day the inference was done, tested, committed his team is still waiting for the cloud GPU queue to free up the only thing between them is 128GB of unified memory in a box smaller than your lunch one Spark is a datacenter node, two Sparks is your own cluster bookmark this and send it to whoever is still paying for cloud GPUs

THIS AI DEVELOPER RUNS 120B PARAMETER MODELS LOCALLY AND HIS MACBOOK PRO CAN’T EVEN OPEN THEM he bought a DGX Spark, gave it the model, went back to his desk by end of day the inference was done, tested, committed his team is still waiting for the cloud GPU queue to free up the only thing between them is 128GB of unified memory in a box smaller than your lunch one Spark is a datacenter node, two Sparks is your own cluster bookmark this and send it to whoever is still paying for cloud GPUs

leopardracer

85,838 views • 5 months ago

i'm running a 397 billion parameter model on a amd ai max box that sits on my desk and pulls less power than a gaming laptop. the model is Nex-N2-Pro, 397B-A17B, the open weight release people are putting next to gpt-5.5 on coding. i have it quantized to IQ1_M, 1.75 bits per weight, 90gb of weights loaded into the 128gb of unified memory on amd's strix halo igpu. watch the gpu in this recording. it spikes, it sustains, it does not fall over. that is the part the spec sheets never show you, not just that a 400b model loads, but that an integrated graphics chip holds the load and generates token after token, stable, no crash, no thermal cliff. and it is not a slideshow. roughly 18 tokens a second, faster than you can read. a frontier scale model producing usable output, fully local. no datacenter, no rented h100s, no api key, no permission. three years ago a model this size meant a server room and a budget to match. tonight it is a quiet box on my desk. this is the accessible tier almost nobody benchmarks honestly, and it is further along than the timeline thinks. the full breakdown is coming, rocm vs vulkan on this chip, and this little amd box head to head against the nvidia equivalent. stay tuned.

i'm running a 397 billion parameter model on a amd ai max box that sits on my desk and pulls less power than a gaming laptop. the model is Nex-N2-Pro, 397B-A17B, the open weight release people are putting next to gpt-5.5 on coding. i have it quantized to IQ1_M, 1.75 bits per weight, 90gb of weights loaded into the 128gb of unified memory on amd's strix halo igpu. watch the gpu in this recording. it spikes, it sustains, it does not fall over. that is the part the spec sheets never show you, not just that a 400b model loads, but that an integrated graphics chip holds the load and generates token after token, stable, no crash, no thermal cliff. and it is not a slideshow. roughly 18 tokens a second, faster than you can read. a frontier scale model producing usable output, fully local. no datacenter, no rented h100s, no api key, no permission. three years ago a model this size meant a server room and a budget to match. tonight it is a quiet box on my desk. this is the accessible tier almost nobody benchmarks honestly, and it is further along than the timeline thinks. the full breakdown is coming, rocm vs vulkan on this chip, and this little amd box head to head against the nvidia equivalent. stay tuned.

Sudo su

32,163 views • 4 months ago

CHINESE AI TEACHER STACKED 4 NVIDIA DGX SPARK CHIPS INTO A CLUSTER - AND NOW RUNS A BUSINESS ON IT FROM HIS DESK 4 chips at $2,999 each - one time investment - and now he has a cluster that handles client inference requests around the clock while he does something else he's running a Qwen 32 35B model with 4 concurrent requests at the same time and the speed barely drops compared to a single request what used to require a $10,000+/month cloud setup now sits on his desk and costs him $40-60/month in electricity 4 chips at $2,999 each - one time investment - and now he has a cluster that handles client inference requests around the clock while he does something else the people still paying per token are financing someone else's infrastructure

CHINESE AI TEACHER STACKED 4 NVIDIA DGX SPARK CHIPS INTO A CLUSTER - AND NOW RUNS A BUSINESS ON IT FROM HIS DESK 4 chips at $2,999 each - one time investment - and now he has a cluster that handles client inference requests around the clock while he does something else he's running a Qwen 32 35B model with 4 concurrent requests at the same time and the speed barely drops compared to a single request what used to require a $10,000+/month cloud setup now sits on his desk and costs him $40-60/month in electricity 4 chips at $2,999 each - one time investment - and now he has a cluster that handles client inference requests around the clock while he does something else the people still paying per token are financing someone else's infrastructure

Sprytix

51,132 views • 5 months ago

APPLE RESEARCH SCIENTIST JUST SHOWED HOW 4 MAC STUDIOS RUN A TRILLION PARAMETER MODEL LOCALLY ZERO COSTS 13:18 she shows the main thing - connect 4 Mac Studios and you get 1TB of shared memory - exactly enough to run a trillion parameter model right on your desk Apple's library - and four machines start working as one cluster tensor parallelism: every machine holds part of every layer - all process the same token simultaneously - speed increases 3x compared to a single device fine-tuning: one Mac Studio processes 180 tokens per second four together process 600 and not a single byte of data leaves the room one command in the terminal - and a trillion parameter model answers from your desk and runs 24/7 data centers took years to build to run models like this - Apple did it with four Thunderbolt cables

APPLE RESEARCH SCIENTIST JUST SHOWED HOW 4 MAC STUDIOS RUN A TRILLION PARAMETER MODEL LOCALLY ZERO COSTS 13:18 she shows the main thing - connect 4 Mac Studios and you get 1TB of shared memory - exactly enough to run a trillion parameter model right on your desk Apple's library - and four machines start working as one cluster tensor parallelism: every machine holds part of every layer - all process the same token simultaneously - speed increases 3x compared to a single device fine-tuning: one Mac Studio processes 180 tokens per second four together process 600 and not a single byte of data leaves the room one command in the terminal - and a trillion parameter model answers from your desk and runs 24/7 data centers took years to build to run models like this - Apple did it with four Thunderbolt cables

Noisy

48,895 views • 4 months ago

8 DGX Sparks on a desk just ran a frontier model at 400K context. A friend who owns the inference budget at a Berlin AI startup, $52K a month in API bills, sent me this at 2 a.m. with one line: "we're buying the hardware." It starts as a disaster. A mountain of cables burying $32K of NVIDIA silicon, nothing talking to anything. 54 minutes later the whole stack answers as 1 machine. 1TB of unified memory. GLM-5.2 loaded across all 8 nodes. The part that made him rewatch twice sits around the 25-minute mark, where 1 networking fix turns 8 boxes into a cluster. Most people quit exactly there. He paused it 4 times. Screenshotted the interface setup. Forwarded it to his CTO without a caption. The cloud bill was never the moat. The cables were.

8 DGX Sparks on a desk just ran a frontier model at 400K context. A friend who owns the inference budget at a Berlin AI startup, $52K a month in API bills, sent me this at 2 a.m. with one line: "we're buying the hardware." It starts as a disaster. A mountain of cables burying $32K of NVIDIA silicon, nothing talking to anything. 54 minutes later the whole stack answers as 1 machine. 1TB of unified memory. GLM-5.2 loaded across all 8 nodes. The part that made him rewatch twice sits around the 25-minute mark, where 1 networking fix turns 8 boxes into a cluster. Most people quit exactly there. He paused it 4 times. Screenshotted the interface setup. Forwarded it to his CTO without a caption. The cloud bill was never the moat. The cables were.

the lich

56,893 views • 3 months ago

HERMES AGENT NOW RUNS ON AN 8GB LAPTOP GPU JUST AS EASILY AS IT RUNS ON A 128GB MINI PC Nous Research shipped the official Hermes Agent Desktop App this week. Someone pointed it at a local llama server running on an RTX 4060 with 16GB system RAM. The integration took two minutes The model behind it: Gemma 4 26B MoE, QAT quantized, running on 8GB of VRAM. A 60k token prompt held a stable 20 tokens a second, flat, no slowdown as context grew. The flags were nothing exotic, just -cmoe -c 248000 on llama.cpp What that 8GB setup does out of the box: reads and patches its own code, runs it in a terminal, debugs errors, manages GitHub repos, spawns sub-agents for parallel work. Browses the web with vision to debug a UI. Schedules cron jobs in plain language. Connects to Notion, Google Workspace, Linear, and Obsidian to manage tasks on its own That's the same agent layer running on a Minisforum MS-S1 MAX with 128GB of unified memory, 96GB of it to the GPU, holding a 120B model at 56 tokens a second instead of a 26B model at 20. Same software, same tool execution, same zero API key. The only thing that changes between an $800 laptop and a $2,000 mini PC is how big a model you can afford to run underneath it The barrier to running a real autonomous agent locally didn't just drop. It dropped all the way down to hardware most people already own

HERMES AGENT NOW RUNS ON AN 8GB LAPTOP GPU JUST AS EASILY AS IT RUNS ON A 128GB MINI PC Nous Research shipped the official Hermes Agent Desktop App this week. Someone pointed it at a local llama server running on an RTX 4060 with 16GB system RAM. The integration took two minutes The model behind it: Gemma 4 26B MoE, QAT quantized, running on 8GB of VRAM. A 60k token prompt held a stable 20 tokens a second, flat, no slowdown as context grew. The flags were nothing exotic, just -cmoe -c 248000 on llama.cpp What that 8GB setup does out of the box: reads and patches its own code, runs it in a terminal, debugs errors, manages GitHub repos, spawns sub-agents for parallel work. Browses the web with vision to debug a UI. Schedules cron jobs in plain language. Connects to Notion, Google Workspace, Linear, and Obsidian to manage tasks on its own That's the same agent layer running on a Minisforum MS-S1 MAX with 128GB of unified memory, 96GB of it to the GPU, holding a 120B model at 56 tokens a second instead of a 26B model at 20. Same software, same tool execution, same zero API key. The only thing that changes between an $800 laptop and a $2,000 mini PC is how big a model you can afford to run underneath it The barrier to running a real autonomous agent locally didn't just drop. It dropped all the way down to hardware most people already own

NO1ennn

40,079 views • 4 months ago

AMD JUST KILLED THE $200 - A - MONTH AI SUBSCRIPTION - A 65-WATT BOX WITH 128GB OF UNIFIED MEMORY THAT RUNS 70B MODELS LOCALLY. that clip is amd's ryzen ai max+ 395 mini box, held up on stage. the whole pitch is one spec: 128gb of high-speed unified memory. why it matters: cpu, gpu and npu share the same 128gb pool, so a model that used to need a rack of vram just fits. what it actually runs: llama 70b in full precision on the built-in gpu open models up to ~100b parameters, locally, no cloud all of it in a mini pc pulling around 65 watts that's the part that ends the subscription: your prompts, your data, your model, on a box on your desk, not metered by the token. honest line: this runs open models, not gpt-5 or claude opus. for chat, coding help and agents on your own data it's more than enough. for the absolute frontier, the cloud still wins. and it isn't free: the box runs ~$1,500 and up. but that's a few months of a $200 sub, then it's yours forever. no monthly ai bill, no tokens metered, no prompt leaving the room. the uncomfortable part isn't that amd made a fast box. it's that "you need the cloud for ai" was a pricing choice, and unified memory just made it optional. follow RetroChainer for the hardware that turns rent into ownership, and bookmark this before your next ai renewal.

AMD JUST KILLED THE $200 - A - MONTH AI SUBSCRIPTION - A 65-WATT BOX WITH 128GB OF UNIFIED MEMORY THAT RUNS 70B MODELS LOCALLY. that clip is amd's ryzen ai max+ 395 mini box, held up on stage. the whole pitch is one spec: 128gb of high-speed unified memory. why it matters: cpu, gpu and npu share the same 128gb pool, so a model that used to need a rack of vram just fits. what it actually runs: llama 70b in full precision on the built-in gpu open models up to ~100b parameters, locally, no cloud all of it in a mini pc pulling around 65 watts that's the part that ends the subscription: your prompts, your data, your model, on a box on your desk, not metered by the token. honest line: this runs open models, not gpt-5 or claude opus. for chat, coding help and agents on your own data it's more than enough. for the absolute frontier, the cloud still wins. and it isn't free: the box runs ~$1,500 and up. but that's a few months of a $200 sub, then it's yours forever. no monthly ai bill, no tokens metered, no prompt leaving the room. the uncomfortable part isn't that amd made a fast box. it's that "you need the cloud for ai" was a pricing choice, and unified memory just made it optional. follow RetroChainer for the hardware that turns rent into ownership, and bookmark this before your next ai renewal.

RetroChainer

16,995 views • 2 months ago

$FOUR DIFFERENT VENDORS ARE NOW SHIPPING GB10 MINI PCs WITH 128GB UNIFIED MEMORY, AND ONE MICROTIK CRS 804 SWITCH CAN CONNECT UP TO EIGHT OF THEM INTO A 1 TERABYTE LOCAL AI CLUSTER 00:00 he points at the MikroTik CRS 804, "you need some kind of switch that'll handle QSFP56 ports like these", the interconnect that makes the whole cluster possible the GB10 ecosystem is no longer just Nvidia. Dell Pro Max GB10, ASUS Ascent GX10, and MSI Edge Expert all ship the same Grace Blackwell Superchip with 128GB of coherent memory. same silicon, different cases, same 200 gigabit ports on the back the CRS 804 is what connects them at prosumer prices. four 400 gigabit QSFP56 ports on one 1U chassis, breakout cables that split each port into two 200 gigabit lanes. one switch drives eight GB10 units in parallel do the math. eight nodes at 128GB each equals 1024GB of pooled unified memory across the cluster. run vLLM, shard a frontier model across all eight, and inference happens locally on hardware that fits in half a rack the real limiter revealed in the stress test was never throttling. it was interconnect topology, exactly the layer this switch fixes at a fraction of enterprise switch pricing $400 a month for combined chatgpt pro and claude code max hits $4,800 a year per developer. a small team of five running through this cluster pays back inside eight months and never expires the article covers the buying ladder for a single desk. this post is proof of the cluster ladder that starts where the desk one ends save this before the GB10 lineup grows past four vendors and prosumer cluster switches move upmarket$

FOUR DIFFERENT VENDORS ARE NOW SHIPPING GB10 MINI PCs WITH 128GB UNIFIED MEMORY, AND ONE MICROTIK CRS 804 SWITCH CAN CONNECT UP TO EIGHT OF THEM INTO A 1 TERABYTE LOCAL AI CLUSTER 00:00 he points at the MikroTik CRS 804, "you need some kind of switch that'll handle QSFP56 ports like these", the interconnect that makes the whole cluster possible the GB10 ecosystem is no longer just Nvidia. Dell Pro Max GB10, ASUS Ascent GX10, and MSI Edge Expert all ship the same Grace Blackwell Superchip with 128GB of coherent memory. same silicon, different cases, same 200 gigabit ports on the back the CRS 804 is what connects them at prosumer prices. four 400 gigabit QSFP56 ports on one 1U chassis, breakout cables that split each port into two 200 gigabit lanes. one switch drives eight GB10 units in parallel do the math. eight nodes at 128GB each equals 1024GB of pooled unified memory across the cluster. run vLLM, shard a frontier model across all eight, and inference happens locally on hardware that fits in half a rack the real limiter revealed in the stress test was never throttling. it was interconnect topology, exactly the layer this switch fixes at a fraction of enterprise switch pricing $400 a month for combined chatgpt pro and claude code max hits $4,800 a year per developer. a small team of five running through this cluster pays back inside eight months and never expires the article covers the buying ladder for a single desk. this post is proof of the cluster ladder that starts where the desk one ends save this before the GB10 lineup grows past four vendors and prosumer cluster switches move upmarket

NO1ennn

59,725 views • 3 months ago

Your cloud agent runs $10–20 a day. This box runs OpenAI's 120B model for $0. Same workload. It just sits in a closet on AMD silicon. A Minisforum mini PC, Strix Halo, 128GB unified memory. The model lives entirely on-device. 56 tokens a second. No API key. Nothing leaves the network. Cloud only handles the exceptions now. Claude or OpenAI by routing, not by default. Everyone calls this "free." Wrong word. It's not free. It's owned. The bill just stops existing. The cloud was never the only option. It was the default nobody bothered to question.

Your cloud agent runs $10–20 a day. This box runs OpenAI's 120B model for $0. Same workload. It just sits in a closet on AMD silicon. A Minisforum mini PC, Strix Halo, 128GB unified memory. The model lives entirely on-device. 56 tokens a second. No API key. Nothing leaves the network. Cloud only handles the exceptions now. Claude or OpenAI by routing, not by default. Everyone calls this "free." Wrong word. It's not free. It's owned. The bill just stops existing. The cloud was never the only option. It was the default nobody bothered to question.

RetroChainer

13,925 views • 4 months ago

YOU DON'T NEED A SERVER RACK. SIX USED MINI PCS AND A $60 CISCO SWITCH WIRE INTO A HOME CLUSTER ON YOUR DESK. that clip is six tiny mini-pcs stacked and patched into one cisco business 110 switch, red cables, a single power strip. each box is a full x86 computer, the kind offices dump by the pallet. used, they run $30 to $60 each. six of them plus the switch is a real cluster for a couple hundred dollars. what a stack like this runs: proxmox or kubernetes spread across all six nodes pi-hole, a nas, home assistant, docker services, each on its own node a local model or two on the boxes with enough ram the trick isn't raw power, it's separation. one node dies, the cluster keeps going. one service crashes, it doesn't drag the rest down. no server rack, no datacenter noise, no monthly cloud bill. the uncomfortable part isn't that it's cheap. it's that the "real infrastructure" you were told needs a rack and a budget fits on a desk and boots off a power strip. follow RetroChainer for more ways to build real infrastructure for pocket money, and bookmark this before the next office dumps its mini pcs.

YOU DON'T NEED A SERVER RACK. SIX USED MINI PCS AND A $60 CISCO SWITCH WIRE INTO A HOME CLUSTER ON YOUR DESK. that clip is six tiny mini-pcs stacked and patched into one cisco business 110 switch, red cables, a single power strip. each box is a full x86 computer, the kind offices dump by the pallet. used, they run $30 to $60 each. six of them plus the switch is a real cluster for a couple hundred dollars. what a stack like this runs: proxmox or kubernetes spread across all six nodes pi-hole, a nas, home assistant, docker services, each on its own node a local model or two on the boxes with enough ram the trick isn't raw power, it's separation. one node dies, the cluster keeps going. one service crashes, it doesn't drag the rest down. no server rack, no datacenter noise, no monthly cloud bill. the uncomfortable part isn't that it's cheap. it's that the "real infrastructure" you were told needs a rack and a budget fits on a desk and boots off a power strip. follow RetroChainer for more ways to build real infrastructure for pocket money, and bookmark this before the next office dumps its mini pcs.

RetroChainer

95,267 views • 2 months ago

watch anon! 27 billion parameter model dancing at 42 tokens a second on a used rtx 3060ti, that costs 200 dollars. consuming 159 watts. temp 62 degrees. pushing almost 100% on the gpu. 6.8 of 8 gigs used with a 128k context window loaded. and every one of those spikes is the model reasoning and firing tool calls. the model is bonsai 27b at 1 bit. PrismML took qwen 3.6 27b, and crushed it to 3.9 gigabytes, about a single sign bit per weight. it drops onto an 8gb rtx 3060 ti, an ampere card from 2020, and it doesn't just load, it serves, it holds the full 128k window, it runs a hermes agent for half an hour straight, it hits its own failures and recovers, it never once drops the loop. the absurd part is the power meter. that same 27b, in its normal quant 4, wants a 24gb card and nearly 17 gigs of weights. this is it at 3.9, doing agentic work at 160 watts, on the tier people call budget. the numbers are sitting right there on nvtop, not in a slide, not a claim, the meter reading while it works. so i keep landing on one question. the model is open, apache 2.0. the card is old and cheap. the whole thing runs offline. why is almost nobody actually running this. is it the hardware, the setup, or did nobody tell you a 27b fits on the card already in your closet.

watch anon! 27 billion parameter model dancing at 42 tokens a second on a used rtx 3060ti, that costs 200 dollars. consuming 159 watts. temp 62 degrees. pushing almost 100% on the gpu. 6.8 of 8 gigs used with a 128k context window loaded. and every one of those spikes is the model reasoning and firing tool calls. the model is bonsai 27b at 1 bit. PrismML took qwen 3.6 27b, and crushed it to 3.9 gigabytes, about a single sign bit per weight. it drops onto an 8gb rtx 3060 ti, an ampere card from 2020, and it doesn't just load, it serves, it holds the full 128k window, it runs a hermes agent for half an hour straight, it hits its own failures and recovers, it never once drops the loop. the absurd part is the power meter. that same 27b, in its normal quant 4, wants a 24gb card and nearly 17 gigs of weights. this is it at 3.9, doing agentic work at 160 watts, on the tier people call budget. the numbers are sitting right there on nvtop, not in a slide, not a claim, the meter reading while it works. so i keep landing on one question. the model is open, apache 2.0. the card is old and cheap. the whole thing runs offline. why is almost nobody actually running this. is it the hardware, the setup, or did nobody tell you a 27b fits on the card already in your closet.

Sudo su

23,691 views • 3 months ago

Okay, this one is sneaky good. Hermes can now run every subagent on a different model. Advanced Delegation is an official Nous Research plugin. You set up named model routes, like fast and deep, and Hermes sends each piece of work down the route that fits. ▸ A fast, cheap model for lookups and summaries ▸ Your strongest model for the hard thinking ▸ A local model on your own endpoint ▸ Or another Hermes profile, with its own model, skills and SOULmd Every result shows the route and model it actually ran on. And Hermes only sees the names and descriptions you write, so it can't invent a route. One agent, a whole bench of models. Setup in the reply.

Hermes Release Watch

26,677 views • 6 days ago

4 MAC MINIS RUNNING A 671B PARAMETER MODEL AS A CLUSTER No data center, no Cloud, no expensive hardware and not a single API call.. Just 4 Mac minis connected through EXO running DeepSeek v3.1 671b locally and actually fast. The part nobody talks about is that you don’t need one monster machine, you can cluster old computers you already have and split the load between them. The full breakdown of tools that make this possible is in the article below. Save and Read it today ↓

4 MAC MINIS RUNNING A 671B PARAMETER MODEL AS A CLUSTER No data center, no Cloud, no expensive hardware and not a single API call.. Just 4 Mac minis connected through EXO running DeepSeek v3.1 671b locally and actually fast. The part nobody talks about is that you don’t need one monster machine, you can cluster old computers you already have and split the load between them. The full breakdown of tools that make this possible is in the article below. Save and Read it today ↓

slash1s

58,724 views • 4 months ago

watch and let this sit for a second anon, a 35B AI model, building a complete multi-file game by itself on hermes agent. no cloud or api. not one line of code written by a human. running on a DGX spark in 2026. because not long ago this was science fiction. this is Ornith, a new open agentic coding model, running near lossless on a single DGX Spark, a 128GB box that sits on your desk. i hand it one spec and step back. it reasons about the architecture, creates the project structure, writes every module one by one, and here's the part that got me, it checks its own work as it goes. makes a call, reads the result, decides if it's right, corrects itself before moving on. no steering from me. this is what people keep underestimating about local AI. they hear "local" and think weaker, slower, a toy. watch the clip. a 35B at near full precision, planning and self correcting on a single desktop. the open models got genuinely good at agentic work, not benchmark good, actually builds the thing good. sped 5x so you can watch the whole build start to finish. this is where local AI actually is now. not coming soon. here.

watch and let this sit for a second anon, a 35B AI model, building a complete multi-file game by itself on hermes agent. no cloud or api. not one line of code written by a human. running on a DGX spark in 2026. because not long ago this was science fiction. this is Ornith, a new open agentic coding model, running near lossless on a single DGX Spark, a 128GB box that sits on your desk. i hand it one spec and step back. it reasons about the architecture, creates the project structure, writes every module one by one, and here's the part that got me, it checks its own work as it goes. makes a call, reads the result, decides if it's right, corrects itself before moving on. no steering from me. this is what people keep underestimating about local AI. they hear "local" and think weaker, slower, a toy. watch the clip. a 35B at near full precision, planning and self correcting on a single desktop. the open models got genuinely good at agentic work, not benchmark good, actually builds the thing good. sped 5x so you can watch the whole build start to finish. this is where local AI actually is now. not coming soon. here.

Sudo su

30,299 views • 4 months ago

AMD CEO Lisa Su showcased a mini PC that could replace your entire AI stack what's changed: - 128GB of unified memory on a single chip - the CPU and GPU share the same memory space - a 235-billion-parameter model can be loaded entirely locally - no cloud, no subscriptions, no logs an active AI user spends around $420 a month on subscriptions - that's $5,280 a year the EVO-X2 costs $1,800 nstall ollama, pull the model, point claude code at localhost. same interface, nothing leaves the machine, nothing costs per request bookmark this and read the article below

AMD CEO Lisa Su showcased a mini PC that could replace your entire AI stack what's changed: - 128GB of unified memory on a single chip - the CPU and GPU share the same memory space - a 235-billion-parameter model can be loaded entirely locally - no cloud, no subscriptions, no logs an active AI user spends around $420 a month on subscriptions - that's $5,280 a year the EVO-X2 costs $1,800 nstall ollama, pull the model, point claude code at localhost. same interface, nothing leaves the machine, nothing costs per request bookmark this and read the article below

kepo

15,451 views • 4 months ago

Kimi K3 dropped 3 days ago and i already used it to build my own PUBG the model that just took the number one coding spot on earth and it runs on a free tier no team no engine i just fed it free open models and assets that are already public and told it what i wanted it built the entire game itself the shooting the recoil the hit detection all of it came straight out of the model i barely wrote a line what youre watching here is the deathmatch mode and its already live battle royale is next im deep in it and the graphics get reworked right after i know it looks raw right now i wont pretend otherwise but the endgame is a build that runs perfect on phones and the moment its on mobile you drop ads in and it earns on its own the part that still gets me is my whole cost for this was a single subscription

Kimi K3 dropped 3 days ago and i already used it to build my own PUBG the model that just took the number one coding spot on earth and it runs on a free tier no team no engine i just fed it free open models and assets that are already public and told it what i wanted it built the entire game itself the shooting the recoil the hit detection all of it came straight out of the model i barely wrote a line what youre watching here is the deathmatch mode and its already live battle royale is next im deep in it and the graphics get reworked right after i know it looks raw right now i wont pretend otherwise but the endgame is a build that runs perfect on phones and the moment its on mobile you drop ads in and it earns on its own the part that still gets me is my whole cost for this was a single subscription

savip.

68,984 views • 3 months ago

Apple built a large foundation model and fine-tuned it on multiple tasks. But they are doing something very clever: They load a single model in memory and use different adapters to specialize the model on the fly. I recorded a video to show you how to write the code to do the same thing Apple is doing. I explain everything step by step. Here is what I'll show you in the video: 1. We'll load two datasets 2. Then load a large model 3. Then, we'll fine-tune the model on both datasets I'll use LoRA to fine-tune the model. This process creates two small adapters, each specializing in solving one of the datasets. The base model's original parameters will remain unchanged. From here: 4. We'll generate a list of tasks 5. We'll load the correct adapter to solve each task The large model I'm using needs 346 MB of memory, but I only need to load it once. Each adapter is only 2.7 MB. I only need to load the base model once and pair it with any of the fine-tuned adapters. Minimum memory footprint and I can solve multiple tasks. Hope this helps!

Apple built a large foundation model and fine-tuned it on multiple tasks. But they are doing something very clever: They load a single model in memory and use different adapters to specialize the model on the fly. I recorded a video to show you how to write the code to do the same thing Apple is doing. I explain everything step by step. Here is what I'll show you in the video: 1. We'll load two datasets 2. Then load a large model 3. Then, we'll fine-tune the model on both datasets I'll use LoRA to fine-tune the model. This process creates two small adapters, each specializing in solving one of the datasets. The base model's original parameters will remain unchanged. From here: 4. We'll generate a list of tasks 5. We'll load the correct adapter to solve each task The large model I'm using needs 346 MB of memory, but I only need to load it once. Each adapter is only 2.7 MB. I only need to load the base model once and pair it with any of the fine-tuned adapters. Minimum memory footprint and I can solve multiple tasks. Hope this helps!

Santiago

84,747 views • 2 years ago