vllm-project
@vllm_project 54.7K X followers vllm-projectFirst product launched 3y ago
- Public revenue
- Undisclosed
- No public revenue across products
Traffic
- vLLM
- +1 with <5K visits or no data
Products
| Product | Visits · Oct | Actions |
|---|---|---|
| vLLMOSSThe High-Throughput and Memory-Efficient inference and serving engine for LLMs | 790.9K | |
| AgDexProduction AI Agent Stack Matrix | <5K |
Timeline
Nov 12 · 6d ago
vLLM optimizes DeepSeek-V4.1-Flash for Vera Rubin NVL72Product Activity#Milestone#Vs. Incumbent
vLLM announced optimizations for DeepSeek-V4.1-Flash on Vera Rubin NVL72, reaching up to 2.52 million tokens/s/GPU peak throughput and 58× the throughput of GB300 NVL72 at 1.7 million tokens/s/GPU and 387 tokens/s/user on SemiAnalysis AgentX.
vLLM adds day-0 support for Thinking Machines' Inkling-2-SmallProduct Activity#Launch Post
vLLM announced day-0 support for Inkling-2-Small by Thinking Machines, a 500B MoE model supporting text and image inputs with text outputs. Inferact led the integration and optimization alongside Thinking Machines Lab, spanning MoE and attention kernels, vision encoding, and speculative decoding.
vLLM details kernel optimizations and benchmark performanceProduct Activity
vLLM shared performance details including DSpark probabilistic drafting acceptance length of ~3.5 at T=1.0 and agentic multi-turn output TPGS of ~1,070 at 32K context and ~700 at 128K on TP8. It also noted fused NVLink-multicast kernels, sequence-parallel prefill, Gluon decode kernels, and support for LoRA and RL via SkyRL.
vLLM vLLM-Omni supports H3-Image across multiple GPU and NPU platformsProduct Activity
vLLM announced that vLLM-Omni supports H3-Image across NVIDIA data-center and RTX GPUs, AMD, Intel, Moore Threads, and Huawei Ascend NPUs.
Nov 7 · 11d ago
vLLM PyTorch hosts Ascend for PyTorch Meetup in BeijingPromotions
On November 7, PyTorch hosted the Ascend for PyTorch Meetup in Beijing, bringing together maintainers from PyTorch core, TorchNPU, vLLM, and over a hundred developers for technical talks, roundtables, and hands-on sessions.
Questions & evidence
- What products has vllm-project built?
- vllm-project tracks 2 standalone software products on EarlyLore (2 currently live), led by vLLM, AgDex.
- How much website traffic do vllm-project's products get?
- Across tracked standalone domains with web traffic estimates, vllm-project's portfolio reaches 790.9K combined monthly visits in each site's latest observed month. Coverage excludes shared third-party hosts and store listings.
Related indie founder research guides
Evaluate your own SaaS idea →