Join the discussion
Write your take first — we'll ask for email only when you're ready to publish.
- Hacker News
- Yes, and yes, again the only way to compete is to build the best not hide in a corner and once again the rest of the world will go on in AI without the United States if we flub it. Circling the wagons, isn’t the long range answer.by Danox
- This is such an obvious conclusion. To take it a bit further...
Scale matters for these things. If we divide the available chips among 5 competing companies we end up with models that are trained on 1/5 of the resources that they otherwise could've been.
Let the companies take turns training on shared hardware, force them to publish results in the open, and then reward them based on how well the resulting model performs at democratically chosen benchmarks. Meritocracy not monopoly.
Make it about how well you wield the silicon, not how much silicon you wield, and make it a positive-sum game. If the people's data is going in, then the people should benefit from what comes out whether or not they have a subscription.
If capitalism as we know it can't complete, so much the worse for capitalism as we know it.
- Open-weight and OSS are wildly different and the article makes a poor comparison.
What's the incentive for the Chinese labs to continue releasing weights 5 years from now? It's not a stable equilibrium and cannot last.
- The lab spending large sums on research and training does not get the inference revenue to fund those efforts.
- Unlike OSS where a single volunteer can keep a project going, training costs run into the $billions.
- OSS is often a two way street where features and integrations are built that the original author benefits from. Open weight models are largely a one way street because the marginal benefit is so much less than training costs.
In the short term, it means Chinese labs can attract talent and, I suspect, funding from their gov. Similar to every other industry the CCP subsidized to take over.
by cheriot - China has no end of money to support these companies. The reason this equilibrium is unstable is that the autonomous agentic coding aspect of the models has been so successfully improved that it will soon be a threat to China state security.by applicative
- > - Unlike OSS where a single volunteer can keep a project going, training costs run into the $billions.
Just some thought: Wouldn't it make sense to build some kind of volunteer computing project to train the next-generation LLM by volunteers, similar to the BOINC [1] projects or Folding@home [2]?
N.B.: BOINC was particularly famous for SETI@home (completed), Einstein@Home, Rosetta@home and PrimeGrid.
I still remember the time when Einstein@Home was in its heyday, and many people who loved putting together fast PCs contributed sometimes even for the reason of showing off in the statistics [3].
---
[1] https://en.wikipedia.org/wiki/Berkeley_Open_Infrastructure_f...
- In fact more countries should have government funded models. There are some obvious issues in China completely dominating open weights space. Kimi had funding of just $2B and could literally create national security threat. A lot of countries could fund something in the range of few billion for something so important. At the very least US and EU could fund few companies.
- why would any software want to have Kubernetes moment? can't count how devop I know that is confused by itby netdur
- I still don't know what it is tbh. Something for docker?
- Recently my company bought another company, and we kept zero of the original engineers, we just had to run the ghost ship.
We walked in, and it was fine. Because it was all kubernetes and laid out like every other app for the most part.
The kube hate is just sad at this point. You need to know like 15 concepts that are all applied in the same way. It mostly just works.
by honkycat - FTFA: American labs need to release frontier-grade open-weight models under licenses that startups can actually build on.
oh now i see, the Chinese government is funding the training and release of their best models to pressure OpenAI, Anthropic, and others to do the same for competition's sake. I don't buy it, this seems more like a way to get SOTA models RL'd to comply with Chinese government approved information distribution. If I have to trust a black box of answers to questions i would trust one from a US for-profit publicly traded company subject to market forces over one approved, and heavily subsidized, by the Chinese government.
by chasd00 - What tools do we have to countermeasure the state sponsored bias in the Chinese models? Doesn’t seem like a smart plan if individuals can just compensate for the bias.by aliasxneo
- No, it’s consistent with what China is doing in other markets, which is dumping product to drive others out of business.
I had a shower thought on how to counteract this, specifically related to the AI dumping. If China is losing substantial money on every token, why wouldn’t an adversary try to maliciously increase consumption? This strategy is not really viable against physical goods dumping because demand is finite and there are large environmental costs. Software demand is infinite and the environmental costs are quite low compared to the financial cost to make it, even with ultra cheap Chinese tokens
- A "US for-profit publicly traded company" is what brought us the Cambridge analytica scandal. I'd rather not trust them with the next wave of consciousness and opinion shaping technology.
Also, oligopolies aren't famous for being strongly bound by market forces, especially when their decision makers are non ironically being treated as if they were heads of state.
https://www.nytimes.com/2026/06/17/world/europe/g7-summit-ai...
- Is anyone using open weight models for agentic coding?
What is your stack (harness, model) and how much do you pay per month?
How would you compare your experience to a typical subsidized plan like Claude Code + Pro plan?
I’m asking because i keep hearing that open weight models are cheap and efficient - is that really the case in practice?
by thih9 - GLM 5.2 awq4 via Opencode, it was better than corporate's fave Sonnet 4.8 and a but worse than Opus 4.8. overall quite capable of a lot of what the dev team needed. Not as good as Sonnet 5 but also never runs out of tokens.
The flip side is that it takes four H200s to run, and that will only let you cache context for maybe three users.
Fingers crossed our Blackwells show up and Kimi 3 really releases weights, because at some point devs are spending a significant portion of their salary on tokens and it's somehow cheaper to buy these ridiculous DGX servers and rack and run them.
by Y_Y - > I’m asking because i keep hearing that open weight models are cheap and efficient - is that really the case in practice?
I think that's the case for people who compare it to proprietary models paid via API—which I think is irrelevant given the majority of people daily driving AI coding are doing on a subscription plan.
The better analysis then is not about AI coding, since there's no subscription plan for Kimi K3.
Instead, compare the cost of running some agentic _task_ that isn't coding which can only be done via API. Think of all the startups wrapping around ChatGPT and Claude to provide some additional set of tools, context, data and hoping to turn it into a profitable service.
To those companies, which are many, open models are the difference between the math working out today vs. praygeing they can scale fast enough to find profitability.
by airstrike - I use opencode or pi harness with the deepseek api for all my at home coding usages.
Deepseek is at least on par with Sonnet (ghcopilot at work)- I don’t use opus, too spendy and I don’t need that level of ability.
The cost is for me was $5/6 months of use. Not a big user I guess! It’s good though, fast enough and incredibly inexpensive.
Been testing Qwen 3.6 28B on a 5090, and it’s also quite good for “free”.
I mostly do small self serving embedded projects based on esp32, so not very complex.
- I use DeepSeek 4 with the VSCode CoPilot plugin. I pay about $10 month on the pay-as-you-go plan.
It's not as good as the frontier models I use at work, but it's plenty capable for the types of tasks I am using it for.
- I am using GLM-5.2 via Ollama Cloud in the $20/month plan. With the same plan I can get many different API keys that I use to run my OpenWebUI server, my opencode and pi dev sessions. I am usually running 2 to 4 sessions concurrently, and I never hit quota limits. At work I get Claude, and I was getting reports that I was spending $75 per hour of work on Opus.by rglullis
- I'm using Kimi K3 + OpenCode. I pay their API pricing, costs about $5 / hour (and chews through ~10 million tokens / hour) during continuous use when I have one or two sessions running and doing their thing.
Can't comment on how it compares to plans (I really don't like the limitations and general shenanigans I see around plans, so I've never tried them).
It is notably slower than Fable / Opus / Gemini, but also vastly cheaper than their API pricing.
by Scene_Cast2 - I'm using qwen 3.6 35B unsloth 4 bit with my 5950x (128 gb memory) and a 3060 12 gb gpu with a self made harness.
At 10k context I get about 40 tps generation and 500 tps prefill. At 100k context I get about 25 tps generation and 400 tps prefill.
It works, but I often use gpt or claude to make a detailed enumerated plan of what I want to do first, then have qwen follow it.
I'm not sure if it is economical or not, but I have solar on the roof so the power use is not really an issue and I already have the hardware.
The biggest benefit for me is that it's all done locally, and I know the harness is not uploading anything or sending telemetry to someone else.
by tyfon - I don’t know if others would find this useful, but previous did have custom harnesses etc.. but tools have improved so much that I drastically simplified.
That said, even the foundational models fail at the hard parts of my code so I use it opportunistically.
I have reduced down to just using zed, will three locally hosted models.
Qwen 3.6 27b on 1x3090 llama.cpp with 128k context ~50tps
Qwen 3.6 35B-A3B on 1x titan v + 2x1080ti llama.cpp with full context ~30tps
GPT-OSS 120b on pure cpu (slow)
I just use zeds parallel agents, task switching, stopping and fixing the code when a model gets stuck.
This still lets me stay engaged, and to modify code to be maintainable etc…
It gets me 80% there and I use to keep a subscription but often times just using googles AI mode is just as good.
That said I have 30 years of experience and insist on knowing how my code works, so this gets me 80% of the short term benefits while not depending on a 3rd party to keep my code moving forward.
Your mileage will vary and 2*5060ti 16gb cards would get around 100/tps with Qwen 3.6 35B-A3B on cards that are widely available.
To be honest the more modern cloud models are using draft tokens etc… that while they are superior for common coding tasks are degrading with more domain specific tasks.
That is just the cost of the draft model being ~10-20% of the foundation models size, and even the biggest Blackwell GPU is limited to ~250/tps so MoE or draft models are required for scaling performance at the foundational level IMHO.
The hard part is my use case are the OOD or small examples in corpus level, the above hurts there.
A Lamborghini may be nice, but I personally need a minivan more.
by nyrikki - Sadly until china scales production of hardware it really isn’t economical to run this stuff yourself. It is good it exists though to put pressure against the labs.
Honestly imo this is just proof apple will win in the end. Eventually a phone will be able to run a model good enough to do most things and it then is game over.
- > Honestly imo this is just proof apple will win in the end. Eventually a phone will be able to run a model good enough to do most things and it then is game over.
I don't see how these are related.
The accuracy and capabilities of your model are directly related to its size. You need a lot of memory for that.
It will be decades before we get enough useful memory in a phone form factor at a price point people can afford it before something like a frontier model now is useful on the phone.
Now, you can run some models on your phone today.
Either way, Apple is using Google today. That could change, but Google isn't exactly getting out of the TPU business and they've been doing it a long time.
Also, some of you live in a very weird Apple bubble. Apple is not so relevant outside the US.
by esseph - Pre bubble prices (~= “we stop building data centers with subsidized credit / circular loans / hidden debt”), a 128GB halo strix ran for $1400, and 200-ish watts. Four of those in a cluster will run a 1T parameter frontier model:
https://www.amd.com/en/developer/resources/technical-article...
At 7 months of claude code subscription per node, the cluster pays for itself in 28 months. On a 5 year (60 month) depreciation schedule, you can buy two of those clusters for basically break even, so you get two concurrent request streams (each of which can batch, etc).
The next generation hardware has already been announced, and should ship roughly two Moore’s law doublings later. It’s likely its steady state price is <= $1400 USD (2024), and it is faster.
So, once the bubble pops (because the financial machinations eventually will come to an abrupt halt), and the labs stop buying hardware for data centers, local inference will be extremely practical and cheaper than a subscription.
My main question is, when that happens, will UNIX Surplus be selling inference servers for pennies on the dollar (like after the dotcom crash), or are the power requirements too exotic for home use?
by hedora - > it really isn’t economical to run this stuff yourself
Quantised models running overnight go most of the way for non-coding tasks.
- > Sadly until china scales production of hardware it really isn’t economical to run this stuff yourself.
I'm running this stuff at home on my desktop and using it through an app on my phone. 60-140TPS depending on model / use case.
It's more than fast enough to even maintain voice conversation.
by esseph - >> do most things and it then is game over.
For the Hyperscalers...and Oracle...cant wait for the day...
by root-parent - Model-on-Chip is coming. GPU are for general computing but have a huge bottle neck for doing model inference.
Even not being able to significantly update a model that is burned on a chip the performance gains are immense. You also don't need the latest chip fabs to make them drastically reducing the cost.
by sschueller - > The government should use procurement to create demand for portable, interoperable systems rather than permanent dependence on one API vendor.
Now, here is an idea that I have not heard before... and I think there is some merit to this. This is also the sort of thing that a state (looking at you CA, CO, IL, NY) could do, instead of just the federal government.