RunPod
Rent H100, A100, and RTX 4090 GPUs by the hour or deploy serverless AI endpoints in seconds.
gpu-cloudh100serverless
View detailsAI Tools Directory
Explore, compare, and discover vetted AI tools, products, and experiments.
Nebius AI Cloud
European AI cloud infrastructure built with NVIDIA H100/H200 clusters and InfiniBand networking.
gpusupercomputerh100
View details
CoreWeave
The AI hyperscaler powering major frontier labs with massive NVIDIA GPU compute clusters.
cloudgpuhyperscale
View details
vLLM
High-throughput and memory-efficient LLM serving engine featuring PagedAttention algorithm.
servingvllmpagedattention
View details
LM Evaluation Harness
The industry-standard framework for evaluating generative language models on 60+ benchmark tasks.
evaluationsbenchmarkseleutherai
View details
Cerebras Cloud
World's fastest AI inference running on Wafer-Scale Engine chips delivering 1,800+ tokens per second.
inferencespeedhardware
View details
SambaNova Cloud
Ultra-fast inference platform running full 405B models at full 16-bit precision at blazing speeds.
hardwareinferencerdu
View details
Text Generation Inference (TGI)
Rust, Python, and gRPC server for text generation inference used in production at Hugging Face.
huggingfacetgiserving
View details
FastChat
Open platform for training, serving, and evaluating LLM chatbots, powering LMSYS Chatbot Arena.
lmsysevaluationserving
View details
DeepSeek-V3
671-billion parameter open-weights Mixture-of-Experts language model with multi-head latent attention.
deepseekmoeopen-source
View details
Llama 3.3 70B
Meta's flagship open-weights model delivering 405B-level capabilities at 70B inference speed.
llamametaopen-source
View details