Can the NVIDIA RTX 4090 run GLM-5.3?
- FP16
- Q8_0
- Q6_K
- Q5_K_M
- Q4_K_M
- Q3_K_M
- Q2_K
No. Even at Q4_K_M, GLM-5.3 needs about 457.8 GB, but the NVIDIA RTX 4090 has 24 GB usable. You could offload some layers to system RAM, but it will be very slow.
Run it in the cloud
Too big for the NVIDIA RTX 4090? Run GLM-5.3 on a rented GPU instead.
Rent a cloud GPU by the hour. No upfront cost, stop any time. Not sure whether to rent or buy? Compare the cost.
Memory needed by quantization
At 8,192 tokens of context. The black line marks the 24 GB the NVIDIA RTX 4090 can use.
| Quant | Quality | Memory | Fits? | Speed | |
|---|---|---|---|---|---|
| FP16 | Full quality | 1507.3 GB | Too big | — | |
| Q8_0 | Near-lossless | 801.4 GB | Too big | — | |
| Q6_K | Excellent | 618.8 GB | Too big | — | |
| Q5_K_M | Very good | 536.9 GB | Too big | — | |
| Q4_K_M | Good, most popular | 457.8 GB | Too big | — | |
| Q3_K_M | Noticeably degraded | 369.4 GB | Too big | — | |
| Q2_K | Heavily degraded | 283.7 GB | Too big | — |
Measured speed from real RTX 4090 owners
Tokens per second that people measured on their own NVIDIA RTX 4090, next to our estimate.
No one has measured GLM-5.3 on the NVIDIA RTX 4090 yet. The speeds above are estimates. Run the benchmark and yours becomes the first real number on this page.
Have an RTX 4090? Add your real speed
One command times GLM-5.3 in your own Ollama, shows you exactly what it would send, and asks before sending anything. No sign-up, no personal data.
macOS and Linux
curl -fsSL https://gpufit.dev/bench.sh | shWindows (PowerShell)
irm https://gpufit.dev/bench.ps1 | iex
Frequently asked questions
Can NVIDIA RTX 4090 run GLM-5.3?
No. Even at Q4_K_M, GLM-5.3 needs about 457.8 GB, but the NVIDIA RTX 4090 has 24 GB usable. You could offload some layers to system RAM, but it will be very slow.
How much VRAM does GLM-5.3 need?
About 457.8 GB at Q4_K_M with 8,192 tokens of context, and about 1507.3 GB at full FP16 precision.
How fast is GLM-5.3 on the NVIDIA RTX 4090?
It does not fit in the NVIDIA RTX 4090's memory, so most layers would run from system RAM and generation would be very slow.