Pick a GPU

Price $0.05/h, fee $0.005 included. 1 h $0.05 · 24 h $1.20. Includes 50 GiB disk, 100 GiB egress.

3.7 TFLOPS per GPU.

Up —Evidence unavailable: not measured · Latency 702 ms · Up to 30 d

Runs 16 AI apps, like Qwen3, BGE-M3 and Llama 3.1

Price $0.258/h, fee $0.026 included. 1 h $0.258 · 24 h $6.192. Includes 50 GiB disk, 100 GiB egress.

55.8 TFLOPS per GPU.

Up —Evidence unavailable: not measured · Latency 164 ms · Up to 30 d

Runs 20 AI apps, like Qwen3, BGE-M3 and Llama 3.1

Price $0.01/h, fee $0.001 included. 1 h $0.01 · 24 h $0.24. Includes 50 GiB disk, 100 GiB egress.

Up —Evidence unavailable: not measured · Latency 763 ms · Up to 30 d

Runs 16 AI apps, like Qwen3, BGE-M3 and Llama 3.1

Price $0.07/h, fee $0.007 included. 1 h $0.07 · 24 h $1.68. Includes 50 GiB disk, 100 GiB egress.

Community reference: ~24 tokens/s on Llama 2 7B Q4_0 (METAL). Another machine and model; your speed may differ. Source.

Up —Evidence unavailable: not measured · Latency —Evidence unavailable: not measured · Up to 30 d

Runs 13 AI apps, like Qwen3, BGE-M3 and Llama 3.1

Price $0.115/h, fee $0.012 included. 1 h $0.115 · 24 h $2.76. Includes 50 GiB disk, 100 GiB egress.

7.5 TFLOPS per GPU.

Up —Evidence unavailable: not measured · Latency 49 ms · Up to 30 d

Runs 20 AI apps, like Qwen3, BGE-M3 and Llama 3.1

Price $0.15/h, fee $0.015 included. 1 h $0.15 · 24 h $3.60. Includes 50 GiB disk, 100 GiB egress.

Up —Evidence unavailable: not measured · Latency 734 ms · Up to 30 d