# flashinfer-ai/flashinfer

FlashInfer: Kernel Library for LLM Serving

Repository: https://github.com/flashinfer-ai/flashinfer
Canonical: https://ross.abutalabs.com/products/flashinfer
Homepage: https://flashinfer.ai
Language: Python
License: Apache-2.0
License Family: permissive
Topics: gpu, large-large-models, cuda, pytorch, llm-inference, jit, attention, nvidia, distributed-inference, moe
Last push: 2026-08-26T23:28:33+00:00

## Health v2 (maintenance only)
Score: 90/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 85, longevity 81
- inputs: {"age_days": 1139, "days_push": 7, "days_rel": 22, "gap_med": 5, "n_releases_24m": 62}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6252, forks 1340 (observed 2026-08-28T04:09:40.979285+00:00)

## What it is
FlashInfer is a GPU kernel library and kernel generator for LLM inference, providing unified APIs for attention, GEMM, and MoE operations with multiple backends (FlashAttention-2/3, cuDNN, CUTLASS, TensorRT-LLM). It supports paged KV-cache, FP8/FP4 low-precision compute, CUDA Graphs, and GPUs from Turing through Blackwell.

## Use cases
- speed up LLM inference serving on NVIDIA GPUs
- efficient attention kernels with paged KV-cache for dynamic batching
- fused mixture-of-experts kernels for MoE models
- FP8 and FP4 quantized attention and GEMM
- accelerate decode and prefill in vLLM or custom serving engines
- shared-prefix cascade attention for memory-efficient batching
- MLA attention kernels for DeepSeek models

## When to choose
- you need state-of-the-art GPU kernel performance for LLM serving
- you're building or tuning an inference engine like vLLM or SGLang
- you need low-precision (FP8/FP4) attention, GEMM, or MoE on modern NVIDIA GPUs
- you want backend flexibility with automatic hardware-aware kernel selection

## When to avoid
- you're running inference on CPUs or non-NVIDIA accelerators without supported backends
- you just need a simple high-level inference API rather than kernel-level control
- your workload is small-scale where kernel optimization overhead isn't worth it

## Facets
- artifact type: library
- maturity: active
- function: llm-inference, gpu-computing, machine-learning, deep-learning, benchmarking
- domain: large-language-models, deep-learning, gpu-computing, machine-learning, developer-tools
- platform: python
- tags: cuda-kernels, attention, mixture-of-experts, kv-cache, fp8, fp4, jit-compilation, vllm, inference-serving, pytorch, gpu, linux, cuda

## Member repositories
- flashinfer-ai/flashinfer (main) score 90

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:40.979285+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:46:36.285294+00:00, confidence not recorded.
  - readme: https://github.com/flashinfer-ai/flashinfer (fetched 2026-08-28T04:09:40.979285+00:00, sha 16122ce128b4)
  - homepage: https://flashinfer.ai (fetched 2026-08-29T08:43:23.021918+00:00, sha 47c1526e3085)
  - site_page: https://docs.flashinfer.ai/ (fetched 2026-08-29T08:43:23.030955+00:00, sha 01c21b9b7af5)
  - site_page: https://flashinfer.ai/releases (fetched 2026-08-29T08:43:23.034325+00:00, sha 2ef5495fa6e3)
- Data as of 2026-08-30T08:39:29.467469+00:00.