Pricing
Pre-alpha: freeEvery job is measured in work units (WU) before it runs, so you know the size up front. Its price in credits is its work units times the rate of the speed tier you pick. Consumers spend credits and Providers earn them. During the pre-alpha nothing is charged: the counts are there so we can tune the rates.
1 work unit
= one sampling step on one megapixel.
The size of the latent is what counts, so a 1024×1024 image costs four times a 512×512 one, and every video frame counts as an image. A second of audio counts as 0.02 megapixels. Each sampler pass counts at the size it runs at, so an upscaled second pass costs more per step than the first. Steps the GPU reuses from your previous run are free. Failed jobs and jobs that move to another GPU cost nothing extra.
Examples
| Job | Settings | WU |
|---|---|---|
| SDXL image | 1024×1024, 30 steps | 31.5 |
| SD 1.5 image | 512×512, 25 steps | 6.6 |
| SDXL batch of 4 | 1024×1024, 30 steps, ×4 | 126 |
| Wan 2.1 video | 832×480, 33 frames, 30 steps | 395 |
| Stable Audio clip | 47 seconds, 50 steps | 47.0 |
Speed tiers
GpuRent sets the rate of each tier; there is no bidding. The ComfyUI plugin shows the expected time and price for each tier before you queue, from the GPUs online, the queue, and how fast each GPU has rendered the model before.
Economy
Waits behind newer Standard jobs for up to 10 minutes, then runs on the slowest free GPU.
An SDXL image (30 WU): 18.0 credits.
Standard
The normal queue, on a GPU that already has the model loaded where possible.
An SDXL image (30 WU): 30.0 credits.
Priority
Jumps the queue by 10 minutes and runs on the fastest measured GPU for the model.
An SDXL image (30 WU): 45.0 credits.
Models in the catalog
Workflows use these by filename. GPUs that already hold one get its jobs first; others download it on demand.
| Model | Type | Size | Min VRAM | Licence | Availability |
|---|---|---|---|---|---|
| sd15 v1-5-pruned-emaonly-fp16.safetensors |
checkpoint | 2.1 GB | 4.0 GB | creativeml-openrail-m | online |
| sdxl-base-1.0 sd_xl_base_1.0.safetensors |
checkpoint | 6.9 GB | 8.0 GB | openrail++ | online |
| wan2.1-t2v-1.3b wan2.1_t2v_1.3B_fp16.safetensors |
diffusion_model | 2.8 GB | 8.0 GB | apache-2.0 | online |
| wan2.1-t2v-14b-q4km wan2.1-t2v-14b-Q4_K_M.gguf |
diffusion_model | 10.12 GB | 11.0 GB | apache-2.0 | online |
| wan2.1-t2v-14b-q5km wan2.1-t2v-14b-Q5_K_M.gguf |
diffusion_model | 11.26 GB | 15.0 GB | apache-2.0 | on demand |
| wan2.2-t2v-a14b-high-q4km Wan2.2-T2V-A14B-HighNoise-Q4_K_M.gguf |
diffusion_model | 9.65 GB | 11.0 GB | apache-2.0 | on demand |
| wan2.2-t2v-a14b-low-q4km Wan2.2-T2V-A14B-LowNoise-Q4_K_M.gguf |
diffusion_model | 9.65 GB | 11.0 GB | apache-2.0 | on demand |
| wan2.2-t2v-a14b-high-q5km Wan2.2-T2V-A14B-HighNoise-Q5_K_M.gguf |
diffusion_model | 10.79 GB | 15.0 GB | apache-2.0 | on demand |
| wan2.2-t2v-a14b-low-q5km Wan2.2-T2V-A14B-LowNoise-Q5_K_M.gguf |
diffusion_model | 10.79 GB | 15.0 GB | apache-2.0 | on demand |
| wan2.2-t2v-lightx2v-4step-high wan2.2_t2v_lightx2v_4steps_lora_v1.1_high_noise.safetensors |
lora | 1.23 GB | 1.0 GB | apache-2.0 | on demand |
| wan2.2-t2v-lightx2v-4step-low wan2.2_t2v_lightx2v_4steps_lora_v1.1_low_noise.safetensors |
lora | 1.23 GB | 1.0 GB | apache-2.0 | on demand |
| umt5-xxl-fp8 umt5_xxl_fp8_e4m3fn_scaled.safetensors |
text_encoder | 6.7 GB | 6.0 GB | apache-2.0 | online |
| wan2.1-vae wan_2.1_vae.safetensors |
vae | 0.3 GB | 1.0 GB | apache-2.0 | online |
| flux2-klein-4b flux-2-klein-4b.safetensors |
diffusion_model | 7.75 GB | 11.0 GB | apache-2.0 | online |
| qwen3-4b qwen_3_4b.safetensors |
text_encoder | 8.04 GB | 8.0 GB | apache-2.0 | online |
| flux2-vae flux2-vae.safetensors |
vae | 0.34 GB | 1.0 GB | apache-2.0 | online |
| z-image-turbo z_image_turbo_bf16.safetensors |
diffusion_model | 12.31 GB | 15.0 GB | apache-2.0 | on demand |
| flux1-ae ae.safetensors |
vae | 0.34 GB | 1.0 GB | apache-2.0 | on demand |
| qwen-image-2512-q4km qwen-image-2512-Q4_K_M.gguf |
diffusion_model | 13.3 GB | 15.0 GB | apache-2.0 | on demand |
| qwen2.5-vl-7b-fp8 qwen_2.5_vl_7b_fp8_scaled.safetensors |
text_encoder | 9.38 GB | 8.0 GB | apache-2.0 | on demand |
| qwen-image-vae qwen_image_vae.safetensors |
vae | 0.25 GB | 1.0 GB | apache-2.0 | on demand |
| acestep-1.5-turbo acestep_v1.5_turbo.safetensors |
diffusion_model | 4.79 GB | 7.0 GB | apache-2.0 | online |
| qwen-0.6b-ace15 qwen_0.6b_ace15.safetensors |
text_encoder | 1.19 GB | 2.0 GB | apache-2.0 | online |
| qwen-1.7b-ace15 qwen_1.7b_ace15.safetensors |
text_encoder | 3.71 GB | 4.0 GB | apache-2.0 | online |
| ace-1.5-vae ace_1.5_vae.safetensors |
vae | 0.34 GB | 1.0 GB | apache-2.0 | online |
| stable-audio-open-1.0 stable-audio-open-1.0.safetensors |
checkpoint | 4.85 GB | 7.0 GB | stability-ai-community | online |
| t5-base t5-base.safetensors |
text_encoder | 0.89 GB | 1.0 GB | apache-2.0 | online |