diff --git a/.gitmodules b/.gitmodules
new file mode 100644
index 0000000..d65e930
--- /dev/null
+++ b/.gitmodules
@@ -0,0 +1,6 @@
+[submodule "open-webui"]
+	path = open-webui
+	url = https://github.com/open-webui/open-webui.git
+[submodule "ollama-webui"]
+	path = ollama-webui
+	url = https://github.com/Sunwood-ai-labs/ollama-webui.git
diff --git a/Dockerfile.openweb.ui b/Dockerfile.openweb.ui
new file mode 100644
index 0000000..59c7da9
--- /dev/null
+++ b/Dockerfile.openweb.ui
@@ -0,0 +1,137 @@
+# syntax=docker/dockerfile:1
+# Initialize device type args
+# use build args in the docker build commmand with --build-arg="BUILDARG=true"
+ARG USE_CUDA=false
+ARG USE_OLLAMA=false
+# Tested with cu117 for CUDA 11 and cu121 for CUDA 12 (default)
+ARG USE_CUDA_VER=cu121
+# any sentence transformer model; models to use can be found at https://huggingface.co/models?library=sentence-transformers
+# Leaderboard: https://huggingface.co/spaces/mteb/leaderboard 
+# for better performance and multilangauge support use "intfloat/multilingual-e5-large" (~2.5GB) or "intfloat/multilingual-e5-base" (~1.5GB)
+# IMPORTANT: If you change the embedding model (sentence-transformers/all-MiniLM-L6-v2) and vice versa, you aren't able to use RAG Chat with your previous documents loaded in the WebUI! You need to re-embed them.
+ARG USE_EMBEDDING_MODEL=sentence-transformers/all-MiniLM-L6-v2
+ARG USE_RERANKING_MODEL=""
+
+######## WebUI frontend ########
+FROM --platform=$BUILDPLATFORM node:21-alpine3.19 as build
+
+WORKDIR /app
+
+COPY ./ollama-webui/package.json ./ollama-webui/package-lock.json ./
+RUN npm ci
+
+COPY ./ollama-webui/ .
+RUN npm run build
+
+######## WebUI backend ########
+FROM python:3.11-slim-bookworm as base
+
+# Use args
+ARG USE_CUDA
+ARG USE_OLLAMA
+ARG USE_CUDA_VER
+ARG USE_EMBEDDING_MODEL
+ARG USE_RERANKING_MODEL
+
+## Basis ##
+ENV ENV=prod \
+    PORT=8080 \
+    # pass build args to the build
+    USE_OLLAMA_DOCKER=${USE_OLLAMA} \
+    USE_CUDA_DOCKER=${USE_CUDA} \
+    USE_CUDA_DOCKER_VER=${USE_CUDA_VER} \
+    USE_EMBEDDING_MODEL_DOCKER=${USE_EMBEDDING_MODEL} \
+    USE_RERANKING_MODEL_DOCKER=${USE_RERANKING_MODEL}
+
+## Basis URL Config ##
+ENV OLLAMA_BASE_URL="/ollama" \
+    OPENAI_API_BASE_URL=""
+
+## API Key and Security Config ##
+ENV OPENAI_API_KEY="" \
+    WEBUI_SECRET_KEY="" \
+    SCARF_NO_ANALYTICS=true \
+    DO_NOT_TRACK=true \
+    ANONYMIZED_TELEMETRY=false
+
+# Use locally bundled version of the LiteLLM cost map json
+# to avoid repetitive startup connections
+ENV LITELLM_LOCAL_MODEL_COST_MAP="True"
+
+
+#### Other models #########################################################
+## whisper TTS model settings ##
+ENV WHISPER_MODEL="base" \
+    WHISPER_MODEL_DIR="/app/backend/data/cache/whisper/models"
+
+## RAG Embedding model settings ##
+ENV RAG_EMBEDDING_MODEL="$USE_EMBEDDING_MODEL_DOCKER" \
+    RAG_RERANKING_MODEL="$USE_RERANKING_MODEL_DOCKER" \
+    SENTENCE_TRANSFORMERS_HOME="/app/backend/data/cache/embedding/models"
+
+## Hugging Face download cache ##
+ENV HF_HOME="/app/backend/data/cache/embedding/models"
+#### Other models ##########################################################
+
+WORKDIR /app/backend
+
+ENV HOME /root
+RUN mkdir -p $HOME/.cache/chroma
+RUN echo -n 00000000-0000-0000-0000-000000000000 > $HOME/.cache/chroma/telemetry_user_id
+
+RUN if [ "$USE_OLLAMA" = "true" ]; then \
+        apt-get update && \
+        # Install pandoc and netcat
+        apt-get install -y --no-install-recommends pandoc netcat-openbsd && \
+        # for RAG OCR
+        apt-get install -y --no-install-recommends ffmpeg libsm6 libxext6 && \
+        # install helper tools
+        apt-get install -y --no-install-recommends curl && \
+        # install ollama
+        curl -fsSL https://ollama.com/install.sh | sh && \
+        # cleanup
+        rm -rf /var/lib/apt/lists/*; \
+    else \
+        apt-get update && \
+        # Install pandoc and netcat
+        apt-get install -y --no-install-recommends pandoc netcat-openbsd && \
+        # for RAG OCR
+        apt-get install -y --no-install-recommends ffmpeg libsm6 libxext6 && \
+        # cleanup
+        rm -rf /var/lib/apt/lists/*; \
+    fi
+
+# install python dependencies
+COPY ./data/litellm/requirements.txt ./requirements.txt
+
+RUN pip3 install uv && \
+    if [ "$USE_CUDA" = "true" ]; then \
+        # If you use CUDA the whisper and embedding model will be downloaded on first use
+        pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/$USE_CUDA_DOCKER_VER --no-cache-dir && \
+        uv pip install --system -r requirements.txt --no-cache-dir && \
+        python -c "import os; from sentence_transformers import SentenceTransformer; SentenceTransformer(os.environ['RAG_EMBEDDING_MODEL'], device='cpu')" && \
+        python -c "import os; from faster_whisper import WhisperModel; WhisperModel(os.environ['WHISPER_MODEL'], device='cpu', compute_type='int8', download_root=os.environ['WHISPER_MODEL_DIR'])"; \
+    else \
+        pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu --no-cache-dir && \
+        uv pip install --system -r requirements.txt --no-cache-dir && \
+        python -c "import os; from sentence_transformers import SentenceTransformer; SentenceTransformer(os.environ['RAG_EMBEDDING_MODEL'], device='cpu')" && \
+        python -c "import os; from faster_whisper import WhisperModel; WhisperModel(os.environ['WHISPER_MODEL'], device='cpu', compute_type='int8', download_root=os.environ['WHISPER_MODEL_DIR'])"; \
+    fi
+
+
+
+# copy embedding weight from build
+# RUN mkdir -p /root/.cache/chroma/onnx_models/all-MiniLM-L6-v2
+# COPY --from=build /app/onnx /root/.cache/chroma/onnx_models/all-MiniLM-L6-v2/onnx
+
+# copy built frontend files
+COPY --from=build /app/build /app/build
+COPY --from=build /app/CHANGELOG.md /app/CHANGELOG.md
+COPY --from=build /app/package.json /app/package.json
+
+# copy backend files
+COPY ./ollama-webui/backend .
+
+EXPOSE 8080
+
+CMD [ "bash", "start.sh"]
\ No newline at end of file
diff --git a/data/litellm/config.yaml b/data/litellm/config.yaml
index e3f420c..481c443 100644
--- a/data/litellm/config.yaml
+++ b/data/litellm/config.yaml
@@ -25,13 +25,14 @@ model_list:
 - litellm_params:
     api_key: "os.environ/ANTHROPIC_API_KEY"
     model: claude-3-haiku-20240307
-  model_name: claude-3-haiku-20240307
+    max_tokens: 2000
+  model_name: claude-3-haiku-20240307-2k
 
-- litellm_params: 
-    model: anthropic.claude-3-haiku-20240307-v1:0
-    ### [OPTIONAL] SET AWS REGION ###
-    aws_region_name: us-west-2
-  model_name: aws_anthropic.claude-3-haiku-20240307-v1:0
+- litellm_params:
+    api_key: "os.environ/ANTHROPIC_API_KEY"
+    model: claude-3-sonnet-20240229
+    max_tokens: 2000
+  model_name: claude-3-sonnet-20240229-2k
 
 - litellm_params: 
     model: anthropic.claude-3-haiku-20240307-v1:0
@@ -39,14 +40,20 @@ model_list:
     aws_region_name: us-west-2
   model_name: aws_anthropic.claude-3-haiku-20240307-v1:0
 
+# - litellm_params: 
+#     model: anthropic.claude-3-haiku-20240307-v1:0
+#     ### [OPTIONAL] SET AWS REGION ###
+#     aws_region_name: us-west-2
+#   model_name: aws_anthropic.claude-3-haiku-20240307-v1:0
+
 - litellm_params: 
     model: anthropic.claude-3-sonnet-20240229-v1:0
     ### [OPTIONAL] SET AWS REGION ###
     aws_region_name: us-west-2
   model_name: aws_anthropic.claude-3-sonnet-20240229-v1:0
   
-- litellm_params: 
-    model: anthropic.claude-3-sonnet-20240229-v1:0
-    ### [OPTIONAL] SET AWS REGION ###
-    aws_region_name: us-west-2
-  model_name: aws_anthropic.claude-3-sonnet-20240229-v1:0
\ No newline at end of file
+# - litellm_params: 
+#     model: anthropic.claude-3-sonnet-20240229-v1:0
+#     ### [OPTIONAL] SET AWS REGION ###
+#     aws_region_name: us-west-2
+#   model_name: aws_anthropic.claude-3-sonnet-20240229-v1:0
\ No newline at end of file
diff --git a/data/litellm/requirements.txt b/data/litellm/requirements.txt
new file mode 100644
index 0000000..0b04c4c
--- /dev/null
+++ b/data/litellm/requirements.txt
@@ -0,0 +1,62 @@
+fastapi==0.109.2
+uvicorn[standard]==0.22.0
+pydantic==2.7.1
+python-multipart==0.0.9
+
+Flask==3.0.3
+Flask-Cors==4.0.0
+
+python-socketio==5.11.2
+python-jose==3.3.0
+passlib[bcrypt]==1.7.4
+uuid==1.30
+
+requests==2.31.0
+aiohttp==3.9.5
+peewee==3.17.3
+peewee-migrate==1.12.2
+psycopg2-binary==2.9.9
+PyMySQL==1.1.0
+bcrypt==4.1.2
+
+litellm
+litellm[proxy]
+
+boto3==1.34.95
+
+argon2-cffi==23.1.0
+APScheduler==3.10.4
+google-generativeai==0.5.2
+
+langchain==0.1.16
+langchain-community==0.0.34
+langchain-chroma==0.1.0
+
+fake-useragent==1.5.1
+chromadb==0.4.24
+sentence-transformers==2.7.0
+pypdf==4.2.0
+docx2txt==0.8
+unstructured==0.11.8
+Markdown==3.6
+pypandoc==1.13
+pandas==2.2.2
+openpyxl==3.1.2
+pyxlsb==1.0.10
+xlrd==2.0.1
+validators==0.28.1
+
+opencv-python-headless==4.9.0.80
+rapidocr-onnxruntime==1.2.3
+
+fpdf2==2.7.8
+rank-bm25==0.2.2
+
+faster-whisper==1.0.1
+
+PyJWT==2.8.0
+PyJWT[crypto]==2.8.0
+
+black==24.4.2
+langfuse==2.27.3
+youtube-transcript-api
diff --git a/docker-compose.ollama.yml b/docker-compose.ollama.yml
index 6d53286..3586e4a 100644
--- a/docker-compose.ollama.yml
+++ b/docker-compose.ollama.yml
@@ -6,7 +6,7 @@ services:
     ports:
       - "11434:11434" 
     volumes:
-      - ./data/ollama:/root/.ollama
+      - ollama-moa2:/root/.ollama
     env_file:
       - .env
     pull_policy: always
@@ -26,23 +26,40 @@ services:
       context: .
       args:
         OLLAMA_BASE_URL: '/ollama'
-      dockerfile: Dockerfile
-    image: ghcr.io/open-webui/open-webui:${WEBUI_DOCKER_TAG-main}
+      dockerfile: Dockerfile.openweb.ui
+      # dockerfile: Dockerfile
+    image: ghcr.io/open-webui/open-webui:main
     volumes:
-      - open-webui:/app/backend/data
+      - open-webui-moa2:/app/backend/data
       - ./data/litellm/config.yaml:/app/backend/data/litellm/config.yaml
+      - ./:/work
     depends_on:
       - ollama
     ports:
       - ${OPEN_WEBUI_PORT-3000}:8080
-      - 14365:14365
+      # - 14365:14365
     env_file:
       - .env
 
-    extra_hosts:
-      - host.docker.internal:host-gateway
+    # extra_hosts:
+    #   - host.docker.internal:host-gateway
     restart: unless-stopped
 
+  litellm:
+    image: ghcr.io/berriai/litellm:main-latest
+    volumes:  
+      - ./data/litellm/config.yaml:/app/config.yaml
+    ports:
+      - "14365:14365"
+    command:
+      - /bin/sh
+      - -c  
+      - |
+        pip install async_generator
+        litellm --config '/app/config.yaml' --debug --host 0.0.0.0 --port 14365 --num_workers 8
+    entrypoint: []
+    env_file:
+      - .env
 volumes:
-  ollama: {}
-  open-webui: {}
\ No newline at end of file
+  ollama-moa2: {}
+  open-webui-moa2: {}
\ No newline at end of file
diff --git a/ollama-webui b/ollama-webui
new file mode 160000
index 0000000..855f2dc
--- /dev/null
+++ b/ollama-webui
@@ -0,0 +1 @@
+Subproject commit 855f2dc6a3dfa4ff1778e2d6b7729f9d74e4872b