[
  {
    "id": "vllm-docs-1",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "Seamless integration with popular Hugging Face models",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-2",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "Efficient management of attention key and value memory with PagedAttention",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-3",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "Continuous batching of incoming requests, chunked prefill, prefix caching",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-4",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "Quantization: FP8, MXFP8/MXFP4, NVFP4, INT8, INT4, GPTQ/AWQ, GGUF, compressed-tensors, ModelOpt, TorchAO, and more",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-5",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "Speculative decoding including n-gram, suffix, EAGLE, DFlash",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-6",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "Streaming outputs",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-7",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "Generation of structured outputs using xgrammar or guidance",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-8",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "Tool calling and reasoning parsers",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-9",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "OpenAI-compatible API server, plus Anthropic Messages API and gRPC support",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-10",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "Efficient multi-LoRA support for dense and MoE layers",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-11",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "Support for NVIDIA GPUs, AMD GPUs, and x86/ARM/PowerPC CPUs. Additionally, diverse hardware plugins such as Google TPUs, Intel Gaudi, IBM Spyre, Huawei Ascend, Rebellions NPU, Apple Silicon, MetaX GPU, and more.",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-12",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "vLLM seamlessly supports 200+ model architectures on HuggingFace",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-13",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "Tensor, pipeline, data, expert, and context parallelism for distributed inference",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-docs-14",
    "tier": "claimed-docs",
    "url": "https://docs.vllm.ai",
    "excerpt": "Disaggregated prefill, decode, and encode",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-gh-1",
    "tier": "github",
    "url": "https://github.com/vllm-project/vllm",
    "excerpt": "Install vLLM with uv (recommended) or pip:",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-gh-2",
    "tier": "github",
    "url": "https://github.com/vllm-project/vllm",
    "excerpt": "Or build from source for development.",
    "fetchedAt": "2026-08-29T01:24:49.366Z"
  },
  {
    "id": "vllm-comm-1",
    "tier": "community",
    "url": "https://news.ycombinator.com/item?id=36409082",
    "excerpt": "I'm spoiled by 4 bit and unfortunately it doesn't appear to be supported here so this isn't of much use to me, but it's awesome to see people working on the inference speed side of things regardless.",
    "fetchedAt": "2026-08-29T01:36:35.298Z"
  },
  {
    "id": "vllm-comm-2",
    "tier": "community",
    "url": "https://news.ycombinator.com/item?id=36409082",
    "excerpt": "Cool, I prefer the OpenAI-Compatible api. Although this is not very technically difficult, it is really intimate, because it make me feel free to use all ChatGPT applications.",
    "fetchedAt": "2026-08-29T01:36:35.298Z"
  },
  {
    "id": "vllm-comm-3",
    "tier": "community",
    "url": "https://news.ycombinator.com/item?id=44407058",
    "excerpt": "We use vLLM kv cache and continuous batching as a foundation for requests in ScalarLM and also add batching optimizations in a centralized queue and by adding explicit batching support in our client... There is more perf you can squeeze out of vLLM.",
    "fetchedAt": "2026-08-29T01:36:35.298Z"
  },
  {
    "id": "vllm-comm-4",
    "tier": "community",
    "url": "https://news.ycombinator.com/item?id=49202852",
    "excerpt": "vLLM is originally marketed as paged attention, but in hindsight, separating the web server and GPU process, continuous batching, kv caching / chunking, and a huge model library including low precision mattered more.",
    "fetchedAt": "2026-08-29T01:36:35.298Z"
  },
  {
    "id": "vllm-comm-5",
    "tier": "community",
    "url": "https://news.ycombinator.com/item?id=46855447",
    "excerpt": "Great job! This is the kind of project that should exist for every complex system. Systems like vLLM's codebase are massive and hard to follow.",
    "fetchedAt": "2026-08-29T01:36:35.298Z"
  },
  {
    "id": "vllm-probe-1",
    "tier": "probe",
    "url": "https://docs.vllm.ai/llms.txt",
    "excerpt": "PROBE llms.txt: HTTP 404 at https://docs.vllm.ai/llms.txt",
    "fetchedAt": "2026-08-29T01:38:29.193Z"
  }
]