GLM-5.3 is now open-weight

https://twitter.com/Zai_org/status/2093354097122455713

100•jeudesprits•59m ago

Comments

scosman•39m ago

I've been using it more and more. Feels like Opus 4.8, in the best possible way.

jonplackett•35m ago

Can you give ant more details how you are you using it? Which harness / service / what you’re building with it etc?

scosman•33m ago

z.ai coder plan, both in opencode and direct API access. I use it for my side projects like https://github.com/scosman/Biscotti (on-device meeting transcription and summaries).

MaxikCZ•34m ago

> in the best possible way

You implying its better than opus 5?

johnnyApplePRNG•33m ago

I'm starting to think Opus 4.8 is significantly smaller than most people assume.

If it's significantly larger than GLM 5.3 (I've heard some insane guesstimates out there like upwards of 5T params or more), that would prove rather embarrassing for Anthropic.

scosman•29m ago

You can't compare models released 6+ months apart. GLM 5.2 was same architecture as 5.3 and not nearly as good. Takes time to build frontier intelligence and distill down to smaller sizes.

walrus01•17m ago

It's not that GLM5.3 in full precision unquantized is any smaller, it's 141 * 5.4GB files at approx 770GB which is about the same size as 5.2.

jasonjmcghee•27m ago

I hear the argument here, but isn't it possible it has dramatically more knowledge and when you get outside the common cases many of us use it for, it'll have completely different capabilities?

I feel like most benchmarks cluster on a reasonably limited area of human knowledge

re-thc•8m ago

> that would prove rather embarrassing for Anthropic

Not really, in that you just work with different constraints.

Anthropic and US labs in general has maybe 100s to 1000s of GPUs per person to experiment. Zai and Chinese labs in general have 1-10.

The priorities are different.

mlnj•26m ago

I have been only using GLM models since last December and have had the best experience without any drama about tokens and geopolitical restrictions. The quality has been great and I am doing more and more with the latest 5.3 and am really excited that consumer hardware will develop in the next few years where I can run these at home.

m00dy•34m ago

GLM-5.3-Flash is actually cheaper than deepseek and better than deepseek but no one is talking about yet :)

scosman•31m ago

It's actually slightly more expensive ($0.50 vs $0.48), but there's a temporary 50% discount.

I've seen dozens of conversations about it in last 24 hours, and every major inference provided added in first 24 hours. I think it's gaining plenty of traction.

swiftcoder•12m ago

It's interesting that OpenCode Go is treating it as 2x more expensive than DeepSeek Flash, even factoring in the 50% discount

re-thc•6m ago

Go has API pricing + this weird scaling of how much is it worth. Some models get $60 of usage, some $30 and some $15 etc.

esafak•30m ago

It is a slow for me through z.ai; it does not feel 'flash' at all. But then neither did the new DS Flash. I think they were getting hammered.

JoeLee1991•29m ago

I've been using it quite a bit too. My main complaint is that it can be really slow sometimes — like, really slow — and the speed feels pretty inconsistent.

fra•29m ago

h/t to DeepInfra for being the first 3rd party provider for it on OpenRouter (https://openrouter.ai/z-ai/glm-5.3?endpoint=b711bea7-3994-49...).

andrewmunsell•14m ago

It's also now live on Ollama Cloud as of a couple minutes ago

ljlolel•5m ago

on my TrustedRouter:

z-ai/glm-5.3: Z.ai, Novita, Atlas Cloud, IO.NET

revolvingthrow•26m ago

GLM 5.3 is probably the sweet spot open weights model if you want to go beyond deepseek flash or the new glm flash. I used it with pi and had a fairly good time, especially since it’s less touchy about cyber and whatnot than the US guys. It’s slightly behind Kimi in ability but it’s a lot easier to run it, I’d expect prices (and speed!) from third parties to be noticeably better.

Assuming you’re willing to drop a fat stack of cash on the upcoming Mac m5 ultra with 512 gb unified memory, you can even run it locally, quantized to 4 bit. Whether it’s even slightly reasonable, well, my wife would probably skin me alive but maybe yours is more understanding.

walrus01•15m ago

One could also run it locally on a used dual xeon (or amd-equivalent) server with 512GB RAM, albeit slower, if you have a useful workflow for it that's like "take this day's efforts and run it through various analysis agents", combined with giving it one-shot tasks/modules to build overnight. You would want a place like a garage or basement to put the server because it'll be loud.

pal9000i•23m ago

how feasible is it build a SOTA specialized model for some use case e.g. deal sourcing by using this as pre-trained model or a LORA or similar pattern on top? Gonna shoot my shot at a billion dollar business

ChildOfChaos•10m ago

How much usage do you find you get on these kinda models (I know the pricing changes a bit) compared to a $20 sub say for Google AI Pro in anti gravity?

I hate how difficult it is to compare prices when looking at subscriptions.

Would $20 in open router, using models like GLM get me more or less?

nkmnz•10m ago

I'd like to ask Sam Altman if he still thinks that it's too dangerous to publish GPT-3. I mean, no one would use it, but what is his reasoning for not publishing it now, in 2026?

chillfox•9m ago

Seeing the price, I am probably just going to keep using GLM-5.2 until 5.3 gets cheaper.

CVEs in UniFi Security Advisory Bulletin 067

Cherish your physical books and media

Compiling Agent Experience into Persistent Knowledge for Skill Evolution

Proxy Web GUI inspector coming to rama CLI in 0.5

LLMs Can Design Near-Optimal OR Algorithms

Your MCP Client Trusts Too Much: What a Capability Layer Stops

What Remains of the SPR Is Probably Less Useful Than You Think

Show HN: I read 648 banks' exchange boards daily to show what they charge

OpenOffice does not print on Tuesdays (2009)

Show HN: Free lifetime pro access to a crunchbase alternative

BioNTech halts mRNA colorectal cancer vaccine trial after review

Let's fix indecipherable public data with Al

Mobile Games Designed to Be Addictive Get More Kid-Friendly

Japan's 3D Modeling Philosophy [video]

Birding Like It's 1899: Inside a Blockbuster American West Video Game (2019)

AI GPUs probably live longer than three years

Just the rumour of a bug is enough to find an exploit these days

Blue-Collar Jobs Are the New Flashpoint in Data-Center Fight

Nvidia Insists It Can Keep Printing Money to Fund the AI Boom

Investigation of agents' behavior in the OpenAI/HuggingFace hacking incident

Agent ROI, Sales Overriding Roadmap, Forcing Your Disruption

OpenAI Flailed Through a Maze

Telnyx Email API is now generally available

Trump's Working to Give Part of Yosemite to a Private Developer

Proof that your data is worth more than charging for any product

OpenAI gates cyber defense in 44 ChatGPT markets with a 1996 US export list

I Signed Up for Claude Pro, Why I'm Canceling (and What I'm Using Instead)

Show HN: OpenLayer – local Photoshop plugin for ComfyUI (inpaint/outpaint demo)

Qwen 3.8 on 1x32GB Radeon wrote Minecraft clone in 24 hours (Rust, WebGPU, WASM)

Veritas – The AI-Powered Compliance Tracker for Global Startups