Kimi K3.1 Pro VRAM needed and GPU rental cost
Kimi K3.1 Pro has 2780.1 billion parameters (from the model card). Weights alone take 5560.2 GB at FP16. With a 20% allowance for the KV cache and activations, plan for about 6672 GB of VRAM at FP16, 3336 GB at FP8 and 1668 GB at INT4. At FP8 it needs more than 8 GPUs.
- Parameters
- 2780.1B
- From the model card
- FP16
- 6672 GB
- Estimated VRAM
- FP8
- 3336 GB
- Estimated VRAM
- INT4
- 1668 GB
- Estimated VRAM
Memory and cost
Estimated VRAM and the cheapest public rental that fits
Each row is a way to run or tune Kimi K3.1 Pro. The option is the cheapest on-demand price we track, from rows we rate high or medium confidence, on one node of up to 8 GPUs. Change the precision or hours in the calculator.
| Mode | Estimated VRAM | Cheapest public on-demand option | Per month (730 h) | Source |
|---|---|---|---|---|
| Inference, FP16 or BF16 | 6672 GB | Needs more than 8 GPUs of any type, so a multi-node job. | n/a | |
| Inference, FP8 or INT8 | 3336 GB | Needs more than 8 GPUs of any type, so a multi-node job. | n/a | |
| Inference, INT4 (4-bit) | 1668 GB | 6x B300 at Hyperstack, $44.40 per hour | $32,412.00 | Source |
| Fine-tune, LoRA (16-bit base) | 6672 GB | Needs more than 8 GPUs of any type, so a multi-node job. | n/a | |
| Fine-tune, QLoRA (4-bit base) | 1668 GB | 6x B300 at Hyperstack, $44.40 per hour | $32,412.00 | Source |
| Fine-tune, full (Adam, mixed precision) | 53378 GB | Needs more than 8 GPUs of any type, so a multi-node job. | n/a |
Fit by chip
How many GPUs Kimi K3.1 Pro needs, by chip
Number of GPUs whose combined memory covers the estimate. Compare the chips on their price pages.
| Precision | L40S 48 GB | A100 80 GB | H100 80 GB | H200 141 GB | B200 180 GB | MI300X 192 GB | B300 288 GB |
|---|---|---|---|---|---|---|---|
| FP16 (6672 GB) | more than 8 | more than 8 | more than 8 | more than 8 | more than 8 | more than 8 | more than 8 |
| FP8 (3336 GB) | more than 8 | more than 8 | more than 8 | more than 8 | more than 8 | more than 8 | more than 8 |
| INT4 (1668 GB) | more than 8 | more than 8 | more than 8 | more than 8 | more than 8 | more than 8 | 6 |
FAQ
Kimi K3.1 Pro questions
How much VRAM does Kimi K3.1 Pro need?
About 6672 GB at FP16, 3336 GB at FP8 and 1668 GB at INT4. These are estimates: parameters times bytes per parameter, plus 20% for the KV cache and activations. Long contexts and large batches need more.
Can Kimi K3.1 Pro run on one H100?
An H100 has 80 GB. At FP16 it needs more than 8 H100 GPUs, at FP8 more than 8 and at INT4 more than 8, by this estimate.
What does it cost to rent a GPU for Kimi K3.1 Pro?
At INT4 it is 6x B300 at Hyperstack, $44.40/hr. Prices change often, so check the source link on each row. Last checked Nov 10, 2026.
How are these estimates made?
Memory is the parameter count times bytes per parameter (2 for FP16, 1 for FP8, 0.5 for INT4), times 1.2. Fine-tuning uses 2 bytes for LoRA, 0.5 for QLoRA and 16 for full fine-tuning with Adam (2 weights, 2 gradients, 12 optimizer bytes), times 1.2. Real usage depends on the framework, context length and batch size.
Need this capacity?
Tell us the GPU, count, term, region and budget. We reply with the public prices that match and the checks to run before you pay. Free, and we do not sell compute.