deepseek-ai/FlashMLA

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

12872 stars · 1137 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

FlashMLA is DeepSeek's library of optimized CUDA attention kernels implementing Multi-head Latent Attention (MLA), including dense and token-level sparse attention for prefill and decoding stages with FP8 KV cache support. It powers DeepSeek-V3 and DeepSeek-V3.2 models and achieves up to 660 TFLOPS on H800 and 1460 TFLOPS on B200 GPUs.

Use cases

• speed up MLA decoding for DeepSeek-style LLM inference
• run sparse attention with FP8 KV cache on Hopper GPUs
• benchmark attention kernel throughput on H800 or B200
• integrate optimized prefill attention kernels into an inference engine
• serve large language models with memory-efficient attention

When to choose

• you are serving DeepSeek-V3/V3.2 or other MLA-based models on NVIDIA Hopper or Blackwell GPUs
• you need maximum attention throughput for prefill or decoding in a custom inference stack
• you want FP8 KV cache sparse decoding kernels

When to avoid

• you use standard multi-head attention models without MLA support
• you run inference on non-NVIDIA hardware like AMD or Apple GPUs
• you need a turnkey inference server rather than low-level kernels

Facets

library · maturity active

machine-learning llm-inference gpu-computing benchmarking deep-learning large-language-models gpu-computing performance python cpp attention-kernels cuda mla sparse-attention fp8 flash-attention hopper blackwell gpu linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="deepseek-ai/FlashMLA")