---
cited_by:
  - '[[thoughts/inference/deepseek-v4.1-state#31-per-layer-spec-registrations-nvidia-target-model|dsv41 §3.1]]'
  - '[[thoughts/inference/deepseek-v4.1-serving#4-pr-reading-order|dsv41 §4]]'
  - '[[thoughts/inference/deepseek-v4.1-serving#unresolved|dsv41 § unresolved]]'
date: '2026-11-11'
kind: pr
local:
  - file:///Users/aarnphm/workspace/vllm/vllm/v1/attention/ops/rocm_aiter_mla_sparse.py
  - file:///Users/aarnphm/workspace/vllm/vllm/config/speculative.py
  - file:///Users/aarnphm/workspace/vllm/vllm/v1/worker/gpu_model_runner.py
  - file:///Users/aarnphm/workspace/vllm/vllm/v1/core/kv_cache_utils.py
  - file:///Users/aarnphm/workspace/vllm/vllm/v1/simple_kv_offload/manager.py
  - file:///Users/aarnphm/workspace/vllm/vllm/model_executor/layers/quantization/utils/fp8_utils.py
  - file:///Users/aarnphm/workspace/vllm/vllm/model_executor/layers/quantization/__init__.py
merged: '2026-09-11'
number: 56214
source:
  - https://github.com/vllm-project/vllm/blob/e77daef89e18e08321ae7b8b24827eedd5fe8673/vllm/v1/attention/ops/rocm_aiter_mla_sparse.py
  - https://github.com/vllm-project/vllm/blob/e77daef89e18e08321ae7b8b24827eedd5fe8673/vllm/config/speculative.py
  - https://github.com/vllm-project/vllm/blob/e77daef89e18e08321ae7b8b24827eedd5fe8673/vllm/v1/worker/gpu_model_runner.py
  - https://github.com/vllm-project/vllm/blob/e77daef89e18e08321ae7b8b24827eedd5fe8673/vllm/v1/core/kv_cache_utils.py
  - https://github.com/vllm-project/vllm/blob/e77daef89e18e08321ae7b8b24827eedd5fe8673/vllm/v1/simple_kv_offload/manager.py
  - https://github.com/vllm-project/vllm/blob/e77daef89e18e08321ae7b8b24827eedd5fe8673/vllm/config/engram.py
  - https://github.com/vllm-project/vllm/blob/e77daef89e18e08321ae7b8b24827eedd5fe8673/vllm/model_executor/layers/quantization/utils/fp8_utils.py
  - https://github.com/vllm-project/vllm/blob/e77daef89e18e08321ae7b8b24827eedd5fe8673/vllm/model_executor/layers/quantization/__init__.py
state: merged
tags:
  - vllm
  - vllm-pr
  - new-architectures
title: '#56214 [Model] Support DeepSeek-V4.1-Flash'
topic: new-architectures
url: https://github.com/vllm-project/vllm/pull/56214
created: '2026-11-11'
modified: '2026-11-11'
published: '2026-11-11'
pageLayout: default
slug: hinterland/papers/maps/prs/56214
permalink: https://aarnphm.xyz/hinterland/papers/maps/prs/56214.md
generator:
  quartz: v4.6.0
  hostedProvider: Cloudflare
  baseUrl: aarnphm.xyz
full: https://aarnphm.xyz/llms-full.txt
---
Support DeepSeek-V4.1-Flash · full dump: [[thoughts/inference/vllm-prs/56214|#56214]]

