meshive/vllm-benchmark repository overview

⁠meshive/vllm-benchmark

vLLM throughput and latency benchmark harness

⁠About Meshive

Meshive is a task-first GPU cloud for AI training, inference, rendering, and serverless model serving. Bring the workload and we run it on RTX 5090 and RTX PRO GPUs at up to 70% less.

• Website: https://meshive.ai⁠
• Docs: https://docs.meshive.ai⁠
• Pricing: https://meshive.ai/pricing⁠
• Python SDK & CLI: https://pypi.org/project/meshive/⁠