GLM-5.1
GLM-5.1 is an open-weight language model from Z.ai (Zhipu) with 754B parameters (mixture of experts). At 4-bit it needs about 509.1 GB of memory, more than any single device in our directory has.
Overview
Includes room for an 8k context and runtime overhead. More context needs more memory.
Key information
- Lab
- Z.ai (Zhipu)
- Released
- 3 Apr 2026
- Parameters
- 754B · MoE
- Type
- Language
- Context
- 202,752 tokens
- License
- MIT
- HF downloads, 30 days
- 789,840
- Repository
- zai-org/GLM-5.1
More from Z.ai (Zhipu)
View allWill it run on your hardware?
How we estimate| Estimate or device | Memory |
|---|---|
| Q4 | 509.1 GB |
| Q8 | 890.5 GB |
| BF16 | 1670 GB |
| RTX 5060 Ti 16 GB | 16 GB usable |
| RTX 5090 32 GB | 32 GB usable |
| DGX Spark 128 GB | 108.8 GB usable |
| Mac Studio M5 Ultra 256 GB | 192 GB usable |
Single-device memory estimate. Pick a device to update the On your device column.
36 GB usable memory
| Format | Memory needed | On your device | Example hardware |
|---|---|---|---|
| 4-bit (Q4) | ≈ 509.1 GB | Too big 1414% | No single device in our directory has enough memory |
| 8-bit (Q8) | ≈ 890.5 GB | Too big 2474% | No single device in our directory has enough memory |
| 16-bit (BF16) | ≈ 1670 GB | Too big 4639% | No single device in our directory has enough memory |
Fits means the estimate stays within 90% of usable memory. Tight is above 90% and up to 100%. Too big is above 100%. These are memory checks, not speed or compatibility tests.
Mac allowances are 66% of total RAM up to 36 GB, and 75% above. AI boxes allow 85% because their memory is shared with the operating system.
Where to run it
With GGUF
Choose a file for llama.cpp or a compatible app.
unsloth/GLM-5.1-GGUFFind compatible hardwareThrough an API
Baidu (fp8) is the cheapest provider we found.
$0.8988 / $2.8248
Input / output per 1M tokens
₹85 / ₹266 at ₹94.29 per $1
Blended: $1.3803 (₹130) per 1M. Ranked by a 3:1 mix of input and output tokens.
mlx_lm.generate --model mlx-community/GLM-5.1 \
--prompt "Explain how a transformer works."llama-cli -hf unsloth/GLM-5.1-GGUF \
-c 8192 -p "Explain how a transformer works."Example commands built from the repository ids above. They need a working install of MLX or llama.cpp. We have not run them; check each repository for available files.
| Precision | |||||
|---|---|---|---|---|---|
| BaiduCheapest | $1.3803₹130 | $0.8988₹85 | $2.8248₹266 | 202,752 | FP8 |
| Chutes | $1.40₹132 | $0.90₹85 | $2.90₹273 | 202,752 | FP8 |
| StreamLake | $1.4835₹140 | $0.966₹91 | $3.036₹286 | 200,000 | FP8 |
| SiliconFlow | $1.8275₹172 | $1.19₹112 | $3.74₹353 | 204,800 | FP8 |
| AtlasCloud | $1.935₹182 | $1.26₹119 | $3.96₹373 | 202,752 | FP8 |
| Alibaba | $2.0425₹193 | $1.33₹125 | $4.18₹394 | 202,745 | FP8 |
| Novita | $2.135₹201 | $1.38₹130 | $4.40₹415 | 204,800 | FP8 |
| Friendli | $2.15₹203 | $1.40₹132 | $4.40₹415 | 202,752 | Not listed |
| GMICloud | $2.15₹203 | $1.40₹132 | $4.40₹415 | 202,752 | FP8 |
| Z.AI | $2.15₹203 | $1.40₹132 | $4.40₹415 | 202,752 | FP8 |
| Venice | $2.1521₹203 | $1.4014₹132 | $4.4044₹415 | 200,000 | FP8 |
INR at ₹94.29 per $1. Prices fetched 14 Nov 2026.
More about this model
Frequently asked questions
How much memory does GLM-5.1 need? At an 8k context, about 509.1 GB at 4-bit, 890.5 GB at 8-bit and 1670 GB at 16-bit (BF16). A longer context needs more.
Can I run GLM-5.1 on a Mac? No Mac in our directory has enough usable memory for the 4-bit estimate of 509.1 GB. An MLX build is available: mlx-community/GLM-5.1.
Does the full context fit? GLM-5.1 supports up to 202,752 tokens of context. Our memory estimates use 8k, so long prompts need more memory than shown.
What does it cost through an API? The cheapest provider we found is Baidu, at $0.90 per 1M input tokens and $2.82 per 1M output tokens (₹85 and ₹266 at 94.29 rupees per dollar). Ranked by a 3:1 mix of input and output tokens.
License in plain English
MIT. This license permits commercial use, modification and redistribution, as long as the copyright and license notice are kept. This summary is not legal advice.
OMHQ RigBench
Not yet tested. We show no speed or quality score without a measurement.
Revision history and sources
Last updated . Sources:Hugging Face model card, MLX repository, GGUF repository, OpenRouter.