Z.ai's open-weight MoE multimodal model released 2026-08-26: 320B total parameters with 18B activated, 1M context, MIT licensed. Takes text, images, video and documents, and runs locally on SGLang or vLLM.
0
Quantized builds
29
Benchmark records
Quantized builds
Community and official quantized releases; "linked records" counts only benchmarks tied to that exact release.
No quantized builds cataloged yet.
Benchmark data
29 configurations
| Hardware | Framework | Quant | Decode | VRAM used | Samples | Evidence |
|---|---|---|---|---|---|---|
| NVIDIA RTX 4080 Super | SGLang | Q4_K_M | 3.1 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 4080 Super | SGLang | IQ3_S | 4.7 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 4080 Super | SGLang | IQ4_XS | 3.5 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 4080 Super | vLLM | Q4_K_M | 3.2 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 4080 Super | vLLM | IQ3_S | 4.8 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 4080 Super | vLLM | IQ4_XS | 3.6 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5070 Ti | SGLang | Q4_K_M | 3.8 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5070 Ti | SGLang | IQ3_S | 5.7 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5070 Ti | SGLang | IQ4_XS | 4.3 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5070 Ti | vLLM | Q4_K_M | 3.9 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5070 Ti | vLLM | IQ3_S | 5.9 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5070 Ti | vLLM | IQ4_XS | 4.4 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5090 | SGLang | Q4_K_M | 7.6 tok/s | 32 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5090 | SGLang | IQ4_XS | 8.5 tok/s | 32 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5090 | vLLM | Q4_K_M | 7.8 tok/s | 32 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5090 | vLLM | IQ3_S | 11.8 tok/s | 32 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5090 | vLLM | IQ4_XS | 8.8 tok/s | 32 GB | 1 | L0 Self-reported |
| NVIDIA RTX 4090 | SGLang | Q4_K_M | 4.3 tok/s | 24 GB | 1 | L0 Self-reported |
| NVIDIA RTX 4090 | SGLang | IQ3_S | 6.4 tok/s | 24 GB | 1 | L0 Self-reported |
| NVIDIA RTX 4090 | SGLang | IQ4_XS | 4.8 tok/s | 24 GB | 1 | L0 Self-reported |
| NVIDIA RTX 4090 | vLLM | Q4_K_M | 4.4 tok/s | 24 GB | 1 | L0 Self-reported |
| NVIDIA RTX 4090 | vLLM | IQ4_XS | 5 tok/s | 24 GB | 1 | L0 Self-reported |
| NVIDIA RTX 3060 12GB | llama.cpp | IQ2_XXS | 3.1 tok/s | 12 GB | 1 | L0 Self-reported |
| NVIDIA RTX 3090 | llama.cpp | IQ3_S | 5.6 tok/s | 24 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5070 Ti | Strata | IQ2_XS | 8.3 tok/s | 16 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5070 | llama.cpp | IQ2_XS | 4.9 tok/s | 12 GB | 1 | L0 Self-reported |
| NVIDIA RTX 5090 | SGLang | IQ3_S | 11.3 tok/s | 32 GB | 1 | L0 Self-reported |
| NVIDIA RTX 4090 | vLLM | IQ3_S | 6.6 tok/s | 24 GB | 1 | L0 Self-reported |
| 2x NVIDIA RTX 5090 | llama.cpp | UD-IQ4_XS | 22 tok/s | 60.34 GB | 1 | L0 Self-reported |
29 of these records are not linked to a specific quantized build; the full set is on the search page.
Community discussion
No discussions yet — start the first one.
Start a discussion