All models › zai
Arena score1376#178 of 416 in Chat, paid models included
Size107.7 B14.3 B active per token (mixture of experts)
4-bit download65 GBweights only, Q4_K_M
Max context128Ktokens, from the model config
Where to use it for free
- The lab's own chat: chat.z.ai (free with daily limits)
- Download and run it at home: official weights · GGUF files for Ollama and LM Studio
Arena scores
Rank among all models, paid ones included. Source: Arena leaderboard dataset.
Speed on popular hardware
| Hardware | GPU memory | Fits? | Speed |
|---|---|---|---|
| RTX 3060 12 GB | 12 GB | Does not fit | — |
| RTX 4060 Ti 16 GB | 16 GB | Does not fit | — |
| RTX 3090 | 24 GB | GPU + 64 GB RAM | 10 tok/s |
| RTX 4090 | 24 GB | GPU + 64 GB RAM | 11 tok/s |
| RTX 5090 | 32 GB | GPU + 64 GB RAM | 13 tok/s |
| RX 7900 XTX | 24 GB | GPU + 64 GB RAM | 10 tok/s |
| 2 × RTX 3090 | 48 GB | GPU + 64 GB RAM | 16 tok/s |
| 4 × RTX 3090 | 96 GB | Fits in GPU | 41 tok/s |
| Mac M4 Max 128 GB | 96 GB | Fits in GPU | 26 tok/s |
| Ryzen AI Max+ 395 128 GB | 96 GB | Fits in GPU | 12 tok/s |
| 8 × H100 SXM | 640 GB | Fits in GPU | 131 tok/s |
Estimate at 4-bit with 32K context, 64 GB of DDR5 for offload. Speed ≈ memory bandwidth × 70% ÷ data read per token; real numbers vary by about ±30%.
GPU memory needed
| Level | 8K context | 32K context | 128K context | Quality vs full |
|---|---|---|---|---|
| Q3_K_M | 56 GB | 61 GB | 79 GB | +4.7% noticeable loss |
| Q4_K_M | 67 GB | 72 GB | 91 GB | +1.2% very small loss |
| Q5_K_M | 79 GB | 83 GB | 102 GB | +0.4% near-lossless |
| Q6_K | 90 GB | 95 GB | 114 GB | +0.2% near-lossless |
| Q8_0 | 117 GB | 121 GB | 140 GB | ≈ 0% same as full |
| F16 | 218 GB | 222 GB | 241 GB | ≈ 0% same as full |
Weights + context cache (from the model's architecture) + runtime. Quality: increase in perplexity against the full model, from llama.cpp tests, scaled by model size. Simulate your own setup.
Chat score over time
13501400145015001550
MayJunJulAugSeptOct
glm-4.6v 1376Best paid model
Arena score from 17 Apr 2026 to 15 Oct 2026.
Vision score over time
11501200125013001350
MayJunJulAugSeptOct
glm-4.6v 1164Best paid model
Arena score from 10 Apr 2026 to 9 Oct 2026.