diff --git a/.buildkite/scripts/run-prime-rl-test.sh b/.buildkite/scripts/run-prime-rl-test.sh
deleted file mode 100755
index a3f2bf8bf..000000000
--- a/.buildkite/scripts/run-prime-rl-test.sh
+++ /dev/null
@@ -1,64 +0,0 @@
-#!/bin/bash
-# SPDX-License-Identifier: Apache-2.0
-# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
-
-# Setup script for Prime-RL integration tests
-# This script prepares the environment for running Prime-RL tests with nightly vLLM
-
-set -euo pipefail
-
-SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
-REPO_ROOT="$(cd "${SCRIPT_DIR}/../.." && pwd)"
-PRIME_RL_REPO="https://github.com/PrimeIntellect-ai/prime-rl.git"
-PRIME_RL_DIR="${REPO_ROOT}/prime-rl"
-
-if command -v rocm-smi &> /dev/null || command -v rocminfo &> /dev/null; then
-    echo "AMD GPU detected. Prime-RL currently only supports NVIDIA. Skipping..."
-    exit 0
-fi
-
-echo "Setting up Prime-RL integration test environment..."
-
-# Clean up any existing Prime-RL directory
-if [ -d "${PRIME_RL_DIR}" ]; then
-    echo "Removing existing Prime-RL directory..."
-    rm -rf "${PRIME_RL_DIR}"
-fi
-
-# Install UV if not available
-if ! command -v uv &> /dev/null; then
-    echo "Installing UV package manager..."
-    curl -LsSf https://astral.sh/uv/install.sh | sh
-    source "$HOME"/.local/bin/env
-fi
-
-# Clone Prime-RL repository at specific branch for reproducible tests
-PRIME_RL_BRANCH="integ-vllm-main"
-echo "Cloning Prime-RL repository at branch: ${PRIME_RL_BRANCH}..."
-git clone --branch "${PRIME_RL_BRANCH}" --single-branch "${PRIME_RL_REPO}" "${PRIME_RL_DIR}"
-cd "${PRIME_RL_DIR}"
-
-echo "Setting up UV project environment..."
-export UV_PROJECT_ENVIRONMENT=/usr/local
-ln -s /usr/bin/python3 /usr/local/bin/python
-
-# Remove vllm pin from pyproject.toml
-echo "Removing vllm pin from pyproject.toml..."
-sed -i '/vllm==/d' pyproject.toml
-
-# Sync Prime-RL dependencies
-echo "Installing Prime-RL dependencies..."
-uv sync --inexact && uv sync --inexact --all-extras
-
-# Verify installation
-echo "Verifying installations..."
-uv run python -c "import vllm; print(f'vLLM version: {vllm.__version__}')"
-uv run python -c "import prime_rl; print('Prime-RL imported successfully')"
-
-echo "Prime-RL integration test environment setup complete!"
-
-echo "Running Prime-RL integration tests..."
-export WANDB_MODE=offline # this makes this test not require a WANDB_API_KEY
-uv run pytest -vs tests/integration/test_rl.py -m gpu
-
-echo "Prime-RL integration tests completed!"
diff --git a/.buildkite/test-amd.yaml b/.buildkite/test-amd.yaml
index 052c85c22..ba6edb92f 100644
--- a/.buildkite/test-amd.yaml
+++ b/.buildkite/test-amd.yaml
@@ -1622,21 +1622,6 @@ steps:
     - uv pip install --system 'gpt-oss[eval]==0.0.5'
     - VLLM_ROCM_USE_AITER_MHA=0 VLLM_ROCM_USE_AITER=1 VLLM_USE_AITER_UNIFIED_ATTENTION=1 pytest -s -v tests/evals/gpt_oss/test_gpqa_correctness.py --model openai/gpt-oss-20b --metric 0.58
 
-##### RL Integration Tests #####
-- label: Prime-RL Integration Test # 15min
-  mirror_hardwares: [amdexperimental]
-  agent_pool: mi325_2
-  # grade: Blocking
-  timeout_in_minutes: 30
-  optional: true
-  num_gpus: 2
-  working_dir: "/vllm-workspace"
-  source_file_dependencies:
-  - vllm/
-  - .buildkite/scripts/run-prime-rl-test.sh
-  commands:
-    - bash .buildkite/scripts/run-prime-rl-test.sh
-
 ##### EPLB Accuracy Tests #####
 - label: DeepSeek V2-Lite Accuracy
   mirror_hardwares: [amdexperimental, amdproduction]
@@ -3201,20 +3186,6 @@ steps:
     - uv pip install --system 'gpt-oss[eval]==0.0.5'
     - VLLM_ROCM_USE_AITER_MHA=0 VLLM_ROCM_USE_AITER=1 VLLM_USE_AITER_UNIFIED_ATTENTION=1 pytest -s -v tests/evals/gpt_oss/test_gpqa_correctness.py --model openai/gpt-oss-20b --metric 0.58
 
-##### RL Integration Tests #####
-- label: Prime-RL Integration Test # 15min
-  mirror_hardwares: [amdexperimental]
-  agent_pool: mi355_2
-  timeout_in_minutes: 30
-  optional: true
-  num_gpus: 2
-  working_dir: "/vllm-workspace"
-  source_file_dependencies:
-  - vllm/
-  - .buildkite/scripts/run-prime-rl-test.sh
-  commands:
-    - bash .buildkite/scripts/run-prime-rl-test.sh
-
 ##### EPLB Accuracy Tests #####
 - label: DeepSeek V2-Lite Accuracy
   mirror_hardwares: [amdexperimental, amdproduction]
diff --git a/.buildkite/test_areas/e2e_integration.yaml b/.buildkite/test_areas/e2e_integration.yaml
index 958bff5c9..d95b73073 100644
--- a/.buildkite/test_areas/e2e_integration.yaml
+++ b/.buildkite/test_areas/e2e_integration.yaml
@@ -28,16 +28,3 @@ steps:
   working_dir: "/vllm-workspace"
   commands:
   - bash .buildkite/scripts/scheduled_integration_test/qwen30b_a3b_fp8_block_ep_eplb.sh 0.8 200 8020 2 1
-
-- label: Prime-RL Integration (2 GPUs)
-  timeout_in_minutes: 30
-  optional: true
-  soft_fail: true
-  num_devices: 2
-  working_dir: "/vllm-workspace"
-  source_file_dependencies:
-  - vllm/
-  - .buildkite/scripts/run-prime-rl-test.sh
-  commands:
-    - nvidia-smi
-    - bash .buildkite/scripts/run-prime-rl-test.sh