diff --git a/docker/Dockerfile b/docker/Dockerfile
index 61ebf970f..964700e2a 100644
--- a/docker/Dockerfile
+++ b/docker/Dockerfile
@@ -132,9 +132,7 @@ WORKDIR /workspace
 COPY requirements/common.txt requirements/common.txt
 COPY requirements/cuda.txt requirements/cuda.txt
 RUN --mount=type=cache,target=/root/.cache/uv \
-    # TODO: remove apache-tvm-ffi once FlashInfer is fixed https://github.com/flashinfer-ai/flashinfer/issues/1962
-    uv pip install --python /opt/venv/bin/python3 --pre apache-tvm-ffi==0.1.0b15 \
-    && uv pip install --python /opt/venv/bin/python3 -r requirements/cuda.txt \
+    uv pip install --python /opt/venv/bin/python3 -r requirements/cuda.txt \
     --extra-index-url ${PYTORCH_CUDA_INDEX_BASE_URL}/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.')
 
 # cuda arch list used by torch
@@ -356,16 +354,14 @@ RUN --mount=type=cache,target=/root/.cache/uv \
 # Install vllm wheel first, so that torch etc will be installed.
 RUN --mount=type=bind,from=build,src=/workspace/dist,target=/vllm-workspace/dist \
     --mount=type=cache,target=/root/.cache/uv \
-    # TODO: remove apache-tvm-ffi once FlashInfer is fixed https://github.com/flashinfer-ai/flashinfer/issues/1962
-    uv pip install --system --pre apache-tvm-ffi==0.1.0b15 \
-    && uv pip install --system dist/*.whl --verbose \
+    uv pip install --system dist/*.whl --verbose \
         --extra-index-url ${PYTORCH_CUDA_INDEX_BASE_URL}/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.')
 
 # Install FlashInfer pre-compiled kernel cache and binaries
 # https://docs.flashinfer.ai/installation.html
 RUN --mount=type=cache,target=/root/.cache/uv \
-    uv pip install --system flashinfer-cubin==0.4.1 \
-    && uv pip install --system flashinfer-jit-cache==0.4.1 \
+    uv pip install --system flashinfer-cubin==0.5.2 \
+    && uv pip install --system flashinfer-jit-cache==0.5.2 \
         --extra-index-url https://flashinfer.ai/whl/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.') \
     && flashinfer show-config
 
diff --git a/docker/Dockerfile.nightly_torch b/docker/Dockerfile.nightly_torch
index 6dfa56017..b88b9c499 100644
--- a/docker/Dockerfile.nightly_torch
+++ b/docker/Dockerfile.nightly_torch
@@ -246,7 +246,7 @@ RUN pip install setuptools==75.6.0 packaging==23.2 ninja==1.11.1.3 build==1.2.2.
 
 
 # build flashinfer for torch nightly from source around 10 mins
-# release version: v0.4.1
+# release version: v0.5.2
 # todo(elainewy): cache flashinfer build result for faster build
 ENV CCACHE_DIR=/root/.cache/ccache
 RUN --mount=type=cache,target=/root/.cache/ccache \
@@ -254,7 +254,7 @@ RUN --mount=type=cache,target=/root/.cache/ccache \
     echo "git clone flashinfer..." \
     && git clone --recursive https://github.com/flashinfer-ai/flashinfer.git \
     && cd flashinfer \
-    && git checkout v0.4.1\
+    && git checkout v0.5.2 \
     && git submodule update --init --recursive \
     && echo "finish git clone flashinfer..." \
     && rm -rf build \
diff --git a/requirements/cuda.txt b/requirements/cuda.txt
index 5f7d520cd..4e393d6b6 100644
--- a/requirements/cuda.txt
+++ b/requirements/cuda.txt
@@ -12,4 +12,4 @@ torchvision==0.24.0 # Required for phi3v processor. See https://github.com/pytor
 # Build from https://github.com/facebookresearch/xformers/releases/tag/v0.0.32.post1
 xformers==0.0.33+5d4b92a5.d20251029; platform_system == 'Linux' and platform_machine == 'x86_64'  # Requires PyTorch >= 2.9
 # FlashInfer should be updated together with the Dockerfile
-flashinfer-python==0.4.1
+flashinfer-python==0.5.2
diff --git a/tests/kernels/attention/test_flashinfer_trtllm_attention.py b/tests/kernels/attention/test_flashinfer_trtllm_attention.py
index 79981009c..693b849eb 100644
--- a/tests/kernels/attention/test_flashinfer_trtllm_attention.py
+++ b/tests/kernels/attention/test_flashinfer_trtllm_attention.py
@@ -238,9 +238,11 @@ def test_flashinfer_trtllm_decode_with_baseline(
     if q_quant_dtype == FP8_DTYPE and o_quant_dtype == FP4_DTYPE:
         rtol, atol = 7e-2, 9e-2
     elif q_quant_dtype == FP8_DTYPE and o_quant_dtype == FP8_DTYPE:
-        rtol, atol = 2e-2, 4e-2
+        rtol, atol = 3e-2, 4e-2
     elif q_quant_dtype == FP8_DTYPE and o_quant_dtype == dtype:
-        rtol, atol = 1e-2, 2e-2
+        rtol, atol = 2e-2, 2e-2
+    elif kv_quant_dtype == FP8_DTYPE:
+        rtol, atol = 4e-2, 6e-2
     else:
         rtol, atol = 1e-2, 1e-2