diff --git a/.circleci/config.yml b/.circleci/config.yml
index eabf4c6129..884ecbee0d 100644
--- a/.circleci/config.yml
+++ b/.circleci/config.yml
@@ -563,39 +563,6 @@ jobs:
           paths:
             - realtime_translation_coverage.xml
             - realtime_translation_coverage
-  mcp_testing:
-    docker:
-      - *python312_image
-    working_directory: ~/project
-
-    steps:
-      - checkout
-      - setup_google_dns
-      - install_uv
-      - run:
-          name: Install Dependencies
-          command: |
-            uv sync --frozen --all-groups --all-extras --python 3.12
-      # Run pytest and generate JUnit XML report
-      - run:
-          name: Run tests
-          command: |
-            uv run --no-sync python -m pytest -vv tests/mcp_tests --cov=litellm --cov-report=xml -x -s -v --junitxml=test-results/junit.xml --durations=5 -n 2
-          no_output_timeout: 15m
-      - run:
-          name: Rename the coverage files
-          command: |
-            mv coverage.xml mcp_coverage.xml
-            mv .coverage mcp_coverage
-
-      # Store test results
-      - store_test_results:
-          path: test-results
-      - persist_to_workspace:
-          root: .
-          paths:
-            - mcp_coverage.xml
-            - mcp_coverage
   agent_testing:
     docker:
       - *python312_image
@@ -794,39 +761,6 @@ jobs:
           paths:
             - search_coverage.xml
             - search_coverage
-  # Split litellm_mapped_tests into parallel jobs
-  litellm_mapped_tests_proxy_part1:
-    docker:
-      - *python312_image
-    working_directory: ~/project
-    resource_class: large
-    steps:
-      - setup_litellm_test_deps
-      - run:
-          name: Run proxy tests part 1 (high-volume directories)
-          command: |
-            uv run --no-sync python -m prisma generate
-            export PYTHONUNBUFFERED=1
-            uv run --no-sync python -m pytest tests/test_litellm/proxy/guardrails tests/test_litellm/proxy/management_endpoints tests/test_litellm/proxy/_experimental tests/test_litellm/proxy/client tests/test_litellm/proxy/auth --junitxml=test-results/junit-proxy-part1.xml --durations=10 -n 4 --maxfail=5 --timeout=60 -vv --log-cli-level=WARNING -r A
-          no_output_timeout: 15m
-      - store_test_results:
-          path: test-results
-  litellm_mapped_tests_proxy_part2:
-    docker:
-      - *python312_image
-    working_directory: ~/project
-    resource_class: large
-    steps:
-      - setup_litellm_test_deps
-      - run:
-          name: Run proxy tests part 2 (all other tests)
-          command: |
-            uv run --no-sync python -m prisma generate
-            export PYTHONUNBUFFERED=1
-            uv run --no-sync python -m pytest tests/test_litellm/proxy --ignore=tests/test_litellm/proxy/guardrails --ignore=tests/test_litellm/proxy/management_endpoints --ignore=tests/test_litellm/proxy/_experimental --ignore=tests/test_litellm/proxy/client --ignore=tests/test_litellm/proxy/auth --junitxml=test-results/junit-proxy-part2.xml --durations=10 -n 4 --maxfail=5 --timeout=120 -vv --log-cli-level=WARNING -r A
-          no_output_timeout: 15m
-      - store_test_results:
-          path: test-results
   litellm_mapped_enterprise_tests:
     docker:
       - *python312_image
@@ -2072,7 +2006,7 @@ jobs:
       - run:
           name: Combine Coverage
           command: |
-            uv tool run --from 'coverage[toml]==7.10.6' coverage combine realtime_translation_coverage ocr_coverage search_coverage mcp_coverage logging_coverage audio_coverage local_testing_part1_coverage local_testing_part2_coverage pass_through_unit_tests_coverage batches_coverage guardrails_coverage redis_caching_coverage
+            uv tool run --from 'coverage[toml]==7.10.6' coverage combine realtime_translation_coverage ocr_coverage search_coverage logging_coverage audio_coverage local_testing_part1_coverage local_testing_part2_coverage pass_through_unit_tests_coverage batches_coverage guardrails_coverage redis_caching_coverage
             uv tool run --from 'coverage[toml]==7.10.6' coverage xml
       - codecov/upload:
           file: ./coverage.xml
@@ -2407,8 +2341,6 @@ workflows:
           filters: *main_branches
       - realtime_translation_testing:
           filters: *main_branches
-      - mcp_testing:
-          filters: *main_branches
       - agent_testing:
           filters: *main_branches
       - guardrails_testing:
@@ -2423,10 +2355,6 @@ workflows:
           filters: *main_branches
       - litellm_mapped_enterprise_tests:
           filters: *main_branches
-      - litellm_mapped_tests_proxy_part1:
-          filters: *main_branches
-      - litellm_mapped_tests_proxy_part2:
-          filters: *main_branches
       - batches_testing:
           filters: *main_branches
       - litellm_utils_testing:
@@ -2444,14 +2372,11 @@ workflows:
       - upload-coverage:
           requires:
             - realtime_translation_testing
-            - mcp_testing
             - agent_testing
             - google_generate_content_endpoint_testing
             - guardrails_testing
             - ocr_testing
             - search_testing
-            - litellm_mapped_tests_proxy_part1
-            - litellm_mapped_tests_proxy_part2
             - litellm_mapped_enterprise_tests
             - batches_testing
             - litellm_utils_testing
diff --git a/.github/workflows/_test-unit-services-base.yml b/.github/workflows/_test-unit-services-base.yml
index 8e0b3568ae..9de3ac3cf5 100644
--- a/.github/workflows/_test-unit-services-base.yml
+++ b/.github/workflows/_test-unit-services-base.yml
@@ -32,6 +32,11 @@ on:
         required: false
         type: boolean
         default: false
+      dist:
+        description: "pytest-xdist distribution mode (loadscope|load|worksteal|loadfile|no)"
+        required: false
+        type: string
+        default: "loadscope"
       artifact-name:
         description: "Unique name for the coverage artifact (must be unique per run)"
         required: false
@@ -124,6 +129,7 @@ jobs:
           MAX_FAILURES: ${{ inputs.max-failures }}
           WORKERS: ${{ inputs.workers }}
           RERUNS: ${{ inputs.reruns }}
+          DIST: ${{ inputs.dist }}
           DATABASE_URL: ${{ inputs.enable-postgres && secrets.DATABASE_URL || '' }}
         run: |
           if [ "${WORKERS}" = "0" ]; then
@@ -143,7 +149,7 @@ jobs:
               -n "${WORKERS}" \
               --reruns "${RERUNS}" \
               --reruns-delay 1 \
-              --dist=loadscope \
+              --dist="${DIST}" \
               --durations=20 \
               --cov=litellm \
               --cov-report=xml:coverage.xml \
diff --git a/.github/workflows/test-unit-proxy-db.yml b/.github/workflows/test-unit-proxy-db.yml
index a631a7c300..f8d5bc265a 100644
--- a/.github/workflows/test-unit-proxy-db.yml
+++ b/.github/workflows/test-unit-proxy-db.yml
@@ -12,6 +12,18 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}
   cancel-in-progress: true
 
+# Semantic matrix: each shard groups tests by concern (auth, server, logging, …)
+# rather than alphabetical letter ranges. Adding a new test file means adding it
+# to whichever group it belongs to, not reshuffling slices.
+#
+# Design targets:
+#   * Every shard runs in <= 7 minutes on a 4-core runner.
+#   * test_key_generate_prisma.py stays serial (workers=0) — it has event-loop
+#     conflicts with the logging worker when run in parallel.
+#   * test_proxy_utils.py runs in its own shard with --dist=worksteal so xdist
+#     spreads its ~64 functions (many parametrized) across workers instead of
+#     pinning the whole file to a single worker (the default --dist=loadscope
+#     behavior for single-file targets).
 jobs:
   proxy-db:
     permissions:
@@ -22,26 +34,111 @@ jobs:
       fail-fast: false
       matrix:
         include:
-          # Key generation tests must NOT run in parallel (event loop conflicts with logging worker)
+          # Must run serially — event-loop conflict with the logging worker.
           - test-group: key-generation
             test-path: "tests/proxy_unit_tests/test_key_generate_prisma.py"
             workers: 0
-            timeout: 30
-          - test-group: auth-checks
-            test-path: "tests/proxy_unit_tests/test_auth_checks.py tests/proxy_unit_tests/test_user_api_key_auth.py"
-            workers: 8
+            dist: loadscope
             timeout: 20
-          # test_proxy_utils.py is large (168+ parametrized tests) — run it on its
-          # own matrix so --dist=loadscope doesn't pin all of it to a single xdist
-          # worker and push the "remaining" group past the job timeout.
+
+          - test-group: auth-and-jwt
+            test-path: >-
+              tests/proxy_unit_tests/test_auth_checks.py
+              tests/proxy_unit_tests/test_user_api_key_auth.py
+              tests/proxy_unit_tests/test_jwt.py
+              tests/proxy_unit_tests/test_jwt_key_mapping.py
+              tests/proxy_unit_tests/test_proxy_custom_auth.py
+              tests/proxy_unit_tests/test_key_generate_dynamodb.py
+              tests/proxy_unit_tests/test_deployed_proxy_keygen.py
+            workers: 8
+            dist: loadscope
+            timeout: 15
+
+          # Own shard, --dist=worksteal so parametrized cases fan out across workers.
           - test-group: proxy-utils
             test-path: "tests/proxy_unit_tests/test_proxy_utils.py"
             workers: 8
-            timeout: 20
-          - test-group: remaining
-            test-path: "tests/proxy_unit_tests --ignore=tests/proxy_unit_tests/test_key_generate_prisma.py --ignore=tests/proxy_unit_tests/test_auth_checks.py --ignore=tests/proxy_unit_tests/test_user_api_key_auth.py --ignore=tests/proxy_unit_tests/test_proxy_utils.py"
+            dist: worksteal
+            timeout: 15
+
+          - test-group: proxy-server
+            test-path: >-
+              tests/proxy_unit_tests/test_proxy_server.py
+              tests/proxy_unit_tests/test_proxy_server_keys.py
+              tests/proxy_unit_tests/test_proxy_server_caching.py
+              tests/proxy_unit_tests/test_proxy_server_langfuse.py
+              tests/proxy_unit_tests/test_proxy_server_spend.py
+              tests/proxy_unit_tests/test_aproxy_startup.py
+              tests/proxy_unit_tests/test_proxy_config_unit_test.py
+              tests/proxy_unit_tests/test_proxy_routes.py
+              tests/proxy_unit_tests/test_proxy_gunicorn.py
+              tests/proxy_unit_tests/test_server_root_path.py
+              tests/proxy_unit_tests/test_proxy_pass_user_config.py
+              tests/proxy_unit_tests/test_proxy_token_counter.py
             workers: 8
-            timeout: 30
+            dist: loadscope
+            timeout: 15
+
+          - test-group: logging-and-callbacks
+            test-path: >-
+              tests/proxy_unit_tests/test_custom_callback_input.py
+              tests/proxy_unit_tests/test_custom_logger_s3_gcs.py
+              tests/proxy_unit_tests/test_proxy_custom_logger.py
+              tests/proxy_unit_tests/test_proxy_reject_logging.py
+              tests/proxy_unit_tests/test_audit_logs_proxy.py
+              tests/proxy_unit_tests/test_search_api_logging.py
+            workers: 8
+            dist: loadscope
+            timeout: 15
+
+          - test-group: db-and-spend
+            test-path: >-
+              tests/proxy_unit_tests/test_prisma_client_backoff_retry.py
+              tests/proxy_unit_tests/test_db_schema_changes.py
+              tests/proxy_unit_tests/test_db_schema_migration.py
+              tests/proxy_unit_tests/test_e2e_pod_lock_manager.py
+              tests/proxy_unit_tests/test_skills_db.py
+              tests/proxy_unit_tests/test_update_daily_tag_spend.py
+              tests/proxy_unit_tests/test_update_spend.py
+              tests/proxy_unit_tests/test_project_endpoints_prisma.py
+              tests/proxy_unit_tests/test_proxy_encrypt_decrypt.py
+            workers: 8
+            dist: loadscope
+            timeout: 15
+
+          - test-group: guardrails-budget-hooks
+            test-path: >-
+              tests/proxy_unit_tests/test_proxy_setting_guardrails.py
+              tests/proxy_unit_tests/test_banned_keyword_list.py
+              tests/proxy_unit_tests/test_default_end_user_budget_simple.py
+              tests/proxy_unit_tests/test_unit_test_max_model_budget_limiter.py
+              tests/proxy_unit_tests/test_zero_cost_model_budget_bypass.py
+              tests/proxy_unit_tests/test_unit_test_proxy_hooks.py
+            workers: 8
+            dist: loadscope
+            timeout: 15
+
+          - test-group: endpoints-and-responses
+            test-path: >-
+              tests/proxy_unit_tests/test_blog_posts_endpoint.py
+              tests/proxy_unit_tests/test_models_fallback_endpoint.py
+              tests/proxy_unit_tests/test_google_endpoint_routing.py
+              tests/proxy_unit_tests/test_google_gemini_proxy_request.py
+              tests/proxy_unit_tests/test_get_favicon.py
+              tests/proxy_unit_tests/test_get_image.py
+              tests/proxy_unit_tests/test_ui_path_detection.py
+              tests/proxy_unit_tests/test_prompt_test_endpoint.py
+              tests/proxy_unit_tests/test_check_batch_cost.py
+              tests/proxy_unit_tests/test_check_responses_cost.py
+              tests/proxy_unit_tests/test_response_polling_handler.py
+              tests/proxy_unit_tests/test_response_polling_pre_call_checks.py
+              tests/proxy_unit_tests/test_realtime_cache.py
+              tests/proxy_unit_tests/test_proxy_exception_mapping.py
+              tests/proxy_unit_tests/test_custom_tokenizer_bug.py
+              tests/proxy_unit_tests/test_model_response_typing
+            workers: 8
+            dist: loadscope
+            timeout: 15
     uses: ./.github/workflows/_test-unit-services-base.yml
     with:
       test-path: ${{ matrix.test-path }}
@@ -49,6 +146,7 @@ jobs:
       reruns: 2
       timeout-minutes: ${{ matrix.timeout }}
       enable-postgres: true
+      dist: ${{ matrix.dist }}
       artifact-name: proxy-db-${{ matrix.test-group }}
     secrets:
       DATABASE_URL: ${{ secrets.DATABASE_URL }}
diff --git a/.github/workflows/test-unit-proxy-endpoints.yml b/.github/workflows/test-unit-proxy-endpoints.yml
index fafc866a3f..1439b2c07f 100644
--- a/.github/workflows/test-unit-proxy-endpoints.yml
+++ b/.github/workflows/test-unit-proxy-endpoints.yml
@@ -36,6 +36,8 @@ jobs:
         tests/test_litellm/proxy/health_endpoints
         tests/test_litellm/proxy/public_endpoints
         tests/test_litellm/proxy/prompts
+        tests/test_litellm/proxy/rag_endpoints
+        tests/test_litellm/proxy/realtime_endpoints
         tests/test_litellm/proxy/ui_crud_endpoints
       workers: 2
       reruns: 2