Hacker Newsnew | past | comments | ask | show | jobs | submit | syntaxing's commentslogin

Every time a major open weight is released, artificial analysis updates to a new one (v5) now. Exaggeration but some of this feels very apt (its also hilarious) https://www.reddit.com/r/LocalLLM/comments/1xkhha2/mr_incred...

    Sorry, this post has been removed by the moderators of r/LocalLLaMA. 

Has been removed, do you happen to have a copy?


There an irony to your statement...since this model seems to be quantized version of Qwen Flash Next + some sort of post quantize training.

Super cool, I’m more curious how this is manufactured. Would love to learn more about that

What does smallest mean? Even smaller than Muse Glimmer which I doubt? Though it is interesting to see US frontier labs trying to push models that outcompete deepseek flash 1017. It’s a good move from the Deepseek federal government ban. Both Inkling 2 small and this seems to try to fill that gap cost vs performance wise. But I like inkling 2 small more due to open weight even though it has a restrictive license.

All thats missing is a MCP server!

I use stable diffusion cpp with a LLM driving it using CLI. It’s been the quickest way for me to run new diffusion models and require near zero skills but the results come out great.

care to go more into your loop that you've setup

Not much too it! Use Hermes, *Claw, Claude cowork, etc and go like "setup this model <huggingface link> using stablediffusion cpp on this computer." Then when its done, in the same session "Use this as the main image (or video) generator moving forward" and it'll change the config

I’m so desperate for home battery using sodium batteries. Safer and lower temperature it can operate at. Low energy density but I don’t care for home battery.

I want to have viable replacements for all the old school lithium ion battery packs so that I won't wonder when the day will come that one of the many we have will experience thermal runaway while nobody is around to deal with it. IMO a world where we no longer worry about spontaneous battery fires can't happen soon enough.

Battery voltages are dependent on chemistry, or, standard electrode potentials of metals used. It can't be configured like way past fractions of volts. And Li has about the highest potential among metals, which is kind of the point of using Li. Which means straight swapping NMC cells with any substitute will cause overcharging. Maybe there are cell configurations that could work in series packs, but it can't be simple swap-ins.

You can have an interposer charge controller. It isn't complicated. Many/most cells in consumer products already have one.

Standard sodium batteries are much more explosive than LiFE

Can they experience a thermal runaway? Can it be triggered by minor deformations or dents?

Yes and no, respectively.

Deformations and dents do not trigger a thermal runaway in sodium batteries.

A sodium battery is MUCH MORE likely to experience a spontaneous thermal runaway due to being at a high SOC.


https://youtu.be/W1ya_ls1zkA?t=209

CATL has demonstrated much better safety for their sodium EV batteries, but home and energy storage packs are more likely to be made of standard 18650 cells like the one in the video.


LFP batteries are quite a bit safer than the LMN or NMC batteries. The sodium batteries are more stable than LFP I think but I'm not sure how much longer and if that matters too much for home battery usage.

There will probably never be a battery that can power your home but can be used without any kind of worry.

The right thing is to place the battery in a way that you don't have to worry abut the fire, not to expect a high-density battery to be perfectly safe.


LFP batteries are pretty safe. Look up some videos of demonstrations. You can cut them in half with a metal knife and nothing happens besides the loss of capacity, and I suppose the knife heats up from the short-circuit current as well.

I don't know what kind of chemistry shenanigans are going on to make that happen, but it's impressive.


Battery safety has very little to do with energy density. High energy density can make the outcomes worse for an unsafe battery, but doesn’t tell you anything about the risk of the fire to begin with.

The reason current Li-ion batteries are unsafe is that they contain a flammable electrolyte. There are batteries with higher energy density that have solid or semi-solid electrolytes. They’re so safe you should probably worry a lot more about driving your car than charging your battery.

Worrying about LFP or So-ion is probably also a waste of time.


A tank of diesel and a generator are a kind of battery that you have to worry about very little.

Are you seriously claiming that fuel depots bring no dire risks? Really?

Diesel is damn hard to set on fire

They are kind of difficult to recharge using solar panels. They also smell bad and are noisy.

Of course, but that was not the point. Evidently there are ways to store energy that are quite safe, so you can't conclude that all forms of storage are inherently unsafe.

Well diesel can catch fire in certain circumstances and can do a huge amount of damage when it does. I'm not sure it's that much safer than some types of batteries...

When a battery/electric car catches fire, the news are heavily amplified. The fire is spectacular, true, and there is no way to put it out but to wait for the car to consume.

But the reality is that the burning rate of ICE cars is a lot higher than electric cars (only electric, that don't carry a heat source along the battery), by far: https://www.motor.com/2026/07/ev-fires-vs-ice-fires-myth-vs-...

Same tune for diesel generators: they catch fire, fumes must be correctly ventilated or you get a flammable atmosphere, rooms where they are installed must be seriously protected against fire hazards (https://global-generators.com/blog/diesel-generator-fire-saf...). But when one catches fire, only local news report it, nobody outside town ever get notice of it, and people tend to think they are 100% fireproof. Meanwhile, a big lithium battery catches fire in California, and the news might even go global.


inherently self-limiting processes are nothing rare or special

There is an aqueous variant of sodium ion batteries that are safer than lithium, but most sodium ion batteries are NOT that variant and are more dangerous than LiFePO4.

Wait what? All data I see about sodium ion is safer than LFP. I’m by no means an expert and would love to see why it’s the other way around

It's definitely not that cut and dry. Here's a paper with a good breakdown of safety comparisons: https://www.sciencedirect.com/science/article/pii/S2352152X2...

Tldr: while sodium thermal runaways are less dangerous than lithium, they are more likely to happen, and in particular charging a sodium battery to 100% SOC makes thermal runaway significantly more likely than in LFP batteries.

It's true that things like puncture or short circuit are much much less dangerous than with lithium. That is a big part of why the increased thermal runaway risk gets overlooked.


Agreed.

The main thing about LFP is people (stupidly) shove them in a basement or enclosed space because they are "safe".

But the "safe" part is that they outgas hydrogen when in thermal runaway, which, combined with enclosed spaces like basements (where people often have gas heaters or ....), is .... a super super bad idea.

Ironically, in this situation, sodium-ion is worse than LFP in almost every conceivable way - the gas is more flammable, toxic, etc.

People who think sodium-ion is "safer" are being sold a bill of goods.


> the gas is more flammable, toxic, etc

There are many alternative materials for the cathodes of sodium-ion batteries and for now the industry has not settled on one of them as being the best.

The safety properties depend on the cathodic material. The vented gas can be toxic only for one class of materials (Prussian blue analogs) and IIRC none of the currently existing commercial sodium-ion batteries uses this kind of cathodes.

For the other cathode types, the vented gas is hydrogen exactly like in almost any kind of battery, and it is equally flammable for all kinds of batteries.

In the article linked above, the conclusion was that for one kind of cathodic materials, which perhaps is the one used in most currently existing sodium-ion batteries, the risk of damaging the battery by overcharging and the risk of thermal runaway (assuming that the battery management system fails to prevent these), are slightly higher than for LFP but lower than for other lithium-ion batteries.

On the other hand, mechanical damage by puncturing the batteries is less dangerous than for any lithium-ion batteries.

These conclusions are not definitive about sodium-ion batteries, because other cathodic materials will behave differently, and there is still a lot of experience to gain with sodium-ion batteries, so it is certain that the cathodic materials used in 2030 will not be the same as today.


There is also the possibility of releasing hydrogen cyanide during a runaway event for some variants too

That can happen only for the sodium-ion batteries that use 1 of at least 5 possible classes of cathodic materials, i.e. Prussian blue analogs, and it seems that none of the currently existing commercial sodium-ion batteries uses this kind of cathode.

Isn't a better/safer solution than Li or Na is for the industry to switch to solid state?

Sodium batteries will only continue to get cheaper. Cost parity now, but probably closer to 1/2 the price in a few years as production ramps up.

I thought the main advantage in comparison with LFP is the lower price(long term) and better sub-zero performance?

Nope. Almost all batteries I see for LFP have heaters for them at too low temps (around freezing 0C/32F). Where as sodium ion can discharge at -25C

Discharging is not the issue with LiPo4 or other lithium tech. They can work up to negative temperatures without any issue. They can go down to -20°C (-4°F).

The problem is charging the battery at negative, as this damages the internal structure. So you need to heat up the battery (aka, you discharge by warming up heating elements), until you hit the right point. And then you can charge up, what will create heat and thus allow for non-damaging charging.

Some even have a fun trick, where if your going to a EV station with your car (end point route on your build in sat nav). They can start to pre-drain the battery if its really too cold, to warm up the batteries even more. So by the time your at the charger, the batteries are on temperature. I forgot what brand did it that way, seen that feature years ago.

What is different is that Sodium-Ion batteries can charge from negative (i think it was -10°C), as they do not suffer from the same damaging effect.


Nearly all EVs now do "battery preconditioning" when on the way to a charger. They would be pretty unusable without, I know one person had a model that didn't have it (this was a few years back) and he could barely pull more than 10kW in winter from public chargers, vs 100kW+ in summer at the exact same chargers.

Tesla do this if they know they're navigating to a charger.

There's a yttrium-containing variety of LFP that's supposed to work better at lower temperature (down to -45 C). But yttrium itself is quite expensive ($35/kg).

Lithium can discharge just fine at -25C without heaters.

Charging it at or around 0C will destroy it extremely rapidly due to electroplating of lithium.


lithium batteries can usually discharge below 0C, it's the charging at sub-zero temperatures that kills them.

LFMP variant of LFP with added magnesium mostly solved this. Presently it is not as long-lasting as LFP so is not suitable for home batteries, but it covers typical lifespan of a car and solves the issues with cold climate.

Aren't sodium higher density (at least in terms of energy per weight) than LiFe?

Maybe you're right in terms of energy per volume. Sodium are very light weight in my experience.


No.

The ratio of the stored charge per mass is 23/7 = about 3.3 times higher for lithium than for sodium (7 and 23 are the atomic masses of lithium and sodium), and the voltage can be higher by around 10% for lithium, leading to an even higher energy per mass ratio.

Nonetheless, both sodium and lithium are only a small fraction of the mass of lithium-ion batteries and sodium-ion batteries, so the difference in stored energy per mass is much smaller than for the pure metals.

In practice that means that today LFP batteries and sodium-ion batteries have about the same energy per mass ratio (and similar prices), with both having a lower specific energy than the more expensive NMC lithium-ion batteries.

For now, sodium-ion batteries are very new, so they are still a little more expensive than LFP, but eventually they will be significantly cheaper.

Already now, they have important advantages, so there are applications for which they are preferable even when still more expensive, e.g. where it must be possible to charge them at very low temperatures or where it must be possible to discharge them completely.


No, sodium batteries are less energy dense.

Honestly, yes, re energy density. I hadn’t seen home batteries in-person until a couple of years ago. I was surprised by how small they could actually be (still with a decent capacity). I’d totally trade some of that for chemistry with a less… catastrophic failure mode, and I don’t even have a big house, even for non-US standards.

Not speaking one way or the other on sodium ion’s properties specifically. I see from sibling comments that it’s not straightforward.


yeah, I am waiting to replace my home battery for the safer sodium option whenever it becomes possible. I keep hoping.

Are you in the US?


I’m hopeful that sodium batteries are going to come down in cost quickly, but if you can find a provider like Base Power or similar where it’s heavily subsidized, there’s a small upfront fee, and they can provide you the economic benefits until sodium batteries reach favorable economic success criteria, that’s a shim until you can replace that with sodium batteries you own outright at a favorable price. You’d have to model what you can get today with a sodium chemistry versus what you can get on offer from one of these battery service providers.

It’s fundamentally a modeling exercise in the forward economic returns curve in the context of your energy needs. Lithium battery lease today as a power purchase agreement (potentially), and then swapping out for a purchased and owned sodium battery pack when the cost hits the right price. It’s always harder to predict the future the further out you attempt to predict, and this technology state of the art is moving very rapidly.


Hot take, Google should focus on their open weight models. I use Qwen 3.8 and it’s been amazing. I would pay for Max but there’s no provider or server here in the states. If they had an amazing Gemma 5, I would seriously consider trying their closed source models. Model comfort and familiarity is a real thing.

I'd also love to see Gemini 4 Flash

I can’t seem to find the cost? The craziness of deepseek flash 1017 is how ridiculously cheap it is for how capable it is.

AA lists it as $0.30/0.03/1 and cost per task is $1.18, about double that of deepseek

It’s about 2-3X more expensive so about the same token efficiency as 1017. I can see this being a good option for the US.

Wow, that's pretty good.

This is a really good move. I noticed this yesterday when I was in private mode in Orion and the Kagi search went through.

> I noticed this yesterday when I was in private mode

How are they going to enforce the 50 search limit? IP address?


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: