GLM-5.2: Frontier Intelligence, Open Weights
twitter.com
GLM-5.2: Frontier Intelligence, Open Weights
1–10 of 20 posts
Re: GLM-5.2: Frontier Intelligence, Open Weights
#2I have already tested this and it is impressive.
Re: GLM-5.2: Frontier Intelligence, Open Weights
#3If I have a fully maxed out MacBook Pro, would it make sense to just switch from Opus 4.8 to this? I've never tried running local models for coding...
Re: GLM-5.2: Frontier Intelligence, Open Weights
#4Blog post with more information including benchmarks: https://z.ai/blog/glm-5.2
Re: GLM-5.2: Frontier Intelligence, Open Weights
#5Given that DeepSwe is one of the very few coding benchmarks worth taking a look at, this achieves rather excellent result at it (not far from opus 4.8).
From looking at the results and my own impression of 5.1 and other models, I think this is the best Chinese coding model by some non-insignificant margin.
Re: GLM-5.2: Frontier Intelligence, Open Weights
#6Given that DeepSwe is one of the very few coding benchmarks worth taking a look at, this achieves rather excellent result at it (not far from opus 4.8). From looking at the results and my own impression of 5.1 and other models, I think this is the best Chinese coding model by some non-insignificant margin.
I've been very pleased with it's performance over the last few days.
It's definitely not near Opus 4.8 level but it's very impressive nonetheless and it does do design extremely well.
Re: GLM-5.2: Frontier Intelligence, Open Weights
#7Given that DeepSwe is one of the very few coding benchmarks worth taking a look at, this achieves rather excellent result at it (not far from opus 4.8). From looking at the results and my own impression of 5.1 and other models, I think this is the best Chinese coding model by some non-insignificant margin.
I've been very pleased with it's performance over the last few days. It's definitely not near Opus 4.8 level but it's very impressive nonetheless and it does do design extremely well.
> it does do design extremely well
Better than Opus?
Re: GLM-5.2: Frontier Intelligence, Open Weights
#8If I have a fully maxed out MacBook Pro, would it make sense to just switch from Opus 4.8 to this? I've never tried running local models for coding...
HuggingFace says this model has 753B parameters, which will need a lot more RAM than a maxed-out MacBook Pro. With 40B active parameters, running from SSD would need patience.
Re: GLM-5.2: Frontier Intelligence, Open Weights
#9If I have a fully maxed out MacBook Pro, would it make sense to just switch from Opus 4.8 to this? I've never tried running local models for coding...
HuggingFace says this model has 753B parameters, which will need a lot more RAM than a maxed-out MacBook Pro. With 40B active parameters, running from SSD would need patience.
I’ve wondered for a while if anyone is working on very wide channel parallel (kind of like RAID 0) SSD for this purpose. Couple that with a tensor processor and that would be interesting.
Re: GLM-5.2: Frontier Intelligence, Open Weights
#10Zhipu AI is founded by a superstar Tsinghua professor, did an IPO in January (Hong Kong stock exchange) hired half it's past research lab and it's stock is >10x since. This is not a "just distill Claude" thing.