SAVRN Model Hub · GLM-5.2
GLM-5.2 GPU Requirements
GLM-5.2 needs about 1.8 TB of GPU memory at 16-bit, 904 GB at 8-bit and 452 GB at 4-bit: its 753.3B parameters plus a 20% working margin. The cheapest setup at 16-bit is 8x MI325X from $16.00 an hour, about $11,680 a month around the clock. None of the 9 accelerators the SAVRN Index prices holds it on one card at 16-bit.
Every Accelerator, Every Precision
How many cards of each accelerator the SAVRN Index prices it takes to hold GLM-5.2, and what that many cards cost an hour at the lowest listed on-demand price. Memory needed: 1.8 TB at 16-bit, 904 GB at 8-bit, 452 GB at 4-bit.
| Accelerator | Memory per card | Lowest price per card | 16-bit | 8-bit | 4-bit |
|---|---|---|---|---|---|
| H100 Voltage Park |
80 GB | $1.99 | More than 8 | More than 8 | 6 cards $11.94/hr |
| H200 GMI Cloud |
141 GB | $2.60 | More than 8 | 7 cards $18.20/hr |
4 cards $10.40/hr |
| B200 Vultr |
180 GB | $3.50 | More than 8 | 6 cards $21.00/hr |
3 cards $10.50/hr |
| GB200 NVL72 GMI Cloud |
186 GB | $8.00 | More than 8 | 5 cards $40.00/hr |
3 cards $24.00/hr |
| MI300X Vultr |
192 GB | $1.85 | More than 8 | 5 cards $9.25/hr |
3 cards $5.55/hr |
| MI325X Vultr |
256 GB | $2.00 | 8 cards $16.00/hr |
4 cards $8.00/hr |
2 cards $4.00/hr |
| B300 Massed Compute |
268 GB | $6.60 | 7 cards $46.20/hr |
4 cards $26.40/hr |
2 cards $13.20/hr |
| GB300 NVL72 Verda |
279 GB | $8.89 | 7 cards $62.23/hr |
4 cards $35.56/hr |
2 cards $17.78/hr |
| MI355X Vultr |
288 GB | $2.59 | 7 cards $18.13/hr |
4 cards $10.36/hr |
2 cards $5.18/hr |
Running It Around the Clock
| Precision | Cheapest setup | Per hour | Per month (730 hours) |
|---|---|---|---|
| 16-bit | 8x MI325X (Vultr) | $16.00 | $11,680 |
| 8-bit | 4x MI325X (Vultr) | $8.00 | $5,840 |
| 4-bit | 2x MI325X (Vultr) | $4.00 | $2,920 |
One copy of the model on rented cards, busy or idle. Serving more users at once takes more copies or more memory for their contexts.
Questions
How much VRAM does GLM-5.2 need?
About 1.8 TB at 16-bit; about 904 GB at 8-bit; about 452 GB at 4-bit: the weights plus 20% for the runtime and a short context. A long context needs more.
What is the cheapest GPU setup to run GLM-5.2?
At 16-bit, 8x MI325X from $16.00 an hour, at the lowest on-demand price the SAVRN Index lists.
Can GLM-5.2 run on a single H100?
No. An H100 has 80 GB, and GLM-5.2 needs 452 GB even at 4-bit, so it takes more than one card.
Memory is the weights at that precision plus 20% for the runtime and a short context. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026. Setups beyond eight cards, one server, are not listed.