Models

Popular models from Hugging Face that GridPup can serve, grouped by family and size, with the GPU memory each one needs and the cheapest machine that fits it right now. Open a family to compare its sizes and versions (the original, quantized and ONNX versions need less GPU memory) and the GPUs for each.

3,133 families · catalog updated 12:56 UTC

• OpenThinker3 by open-thoughtsText generation · 1 model · ↓ 9617B
  Smallest needs19.7 GB GPU
  From$0.16/h2x RTX 3060
• Llama-3_1-Nemotron by nvidiaText generation · 1 model · ↓ 96051B
  Smallest needs140.5 GB GPU
  From$2.25/h8x RTX 3090
• StableBeluga by stabilityaiText generation · 2 models · ↓ 9567B13B
  Smallest needs19.6 GB GPU
  From$0.16/h2x RTX 3060
• stablelm-2 by stabilityaiText generation · 1 model · ↓ 95412B
  Smallest needs30.1 GB GPU
  From$0.44/hRTX 5000 Ada
• Octopus by NexaAIText generation · 1 model · ↓ 947~2.51B
  Smallest needs8 GB GPU
  From

Model details come from Hugging Face and are refreshed daily. GPU memory is GridPup's estimate: the weights, room for an 8,192-token conversation (chat models) and the runtime's own memory. Large models are split across the GPUs of one machine ("4x RTX 4090"). Only machines with a Vast reliability above 95% that download at 400 Mbit/s or more are counted.