agent-cli 0.72.6__tar.gz → 0.73.0__tar.gz

{agent_cli-0.72.6 → agent_cli-0.73.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agent-cli
-Version: 0.72.6
+Version: 0.73.0
 Summary: A suite of AI-powered command-line tools for text correction, audio transcription, and voice assistance.
 Project-URL: Homepage, https://github.com/basnijholt/agent-cli
 Author-email: Bas Nijholt <bas@nijho.lt>
@@ -54,6 +54,7 @@ Requires-Dist: fastapi[standard]; extra == 'memory'
 Requires-Dist: huggingface-hub>=0.20.0; extra == 'memory'
 Requires-Dist: onnxruntime>=1.17.0; extra == 'memory'
 Requires-Dist: openai>=1.0.0; extra == 'memory'
+Requires-Dist: pydantic-ai-slim[google,openai]>=0.1.1; extra == 'memory'
 Requires-Dist: pyyaml>=6.0.0; extra == 'memory'
 Requires-Dist: transformers>=4.30.0; extra == 'memory'
 Requires-Dist: watchfiles>=0.21.0; extra == 'memory'
@@ -70,6 +71,7 @@ Requires-Dist: huggingface-hub>=0.20.0; extra == 'rag'
 Requires-Dist: markitdown[docx,pdf,pptx]>=0.1.3; extra == 'rag'
 Requires-Dist: onnxruntime>=1.17.0; extra == 'rag'
 Requires-Dist: openai>=1.0.0; extra == 'rag'
+Requires-Dist: pydantic-ai-slim[google,openai]>=0.1.1; extra == 'rag'
 Requires-Dist: transformers>=4.30.0; extra == 'rag'
 Requires-Dist: watchfiles>=0.21.0; extra == 'rag'
 Provides-Extra: server
@@ -2064,8 +2066,12 @@ uv tool install "agent-cli[vad]" -p 3.13
 │                                [env var: OPENAI_API_KEY]                               │
 ╰────────────────────────────────────────────────────────────────────────────────────────╯
 ╭─ LLM Configuration ────────────────────────────────────────────────────────────────────╮
-│ --embedding-model        TEXT  Embedding model to use for vectorization.               │
-│                                [default: text-embedding-3-small]                       │
+│ --embedding-base-url        TEXT  Base URL for embedding API. Falls back to            │
+│                                   --openai-base-url if not set. Useful when using      │
+│                                   different providers for chat vs embeddings.          │
+│                                   [env var: EMBEDDING_BASE_URL]                        │
+│ --embedding-model           TEXT  Embedding model to use for vectorization.            │
+│                                   [default: text-embedding-3-small]                    │
 ╰────────────────────────────────────────────────────────────────────────────────────────╯
 ╭─ Server Configuration ─────────────────────────────────────────────────────────────────╮
 │ --host        TEXT     Host/IP to bind API servers to.                                 │
@@ -2227,8 +2233,12 @@ The `memory proxy` command is the core feature—a middleware server that gives
 │                                [env var: OPENAI_API_KEY]                               │
 ╰────────────────────────────────────────────────────────────────────────────────────────╯
 ╭─ LLM Configuration ────────────────────────────────────────────────────────────────────╮
-│ --embedding-model        TEXT  Embedding model to use for vectorization.               │
-│                                [default: text-embedding-3-small]                       │
+│ --embedding-base-url        TEXT  Base URL for embedding API. Falls back to            │
+│                                   --openai-base-url if not set. Useful when using      │
+│                                   different providers for chat vs embeddings.          │
+│                                   [env var: EMBEDDING_BASE_URL]                        │
+│ --embedding-model           TEXT  Embedding model to use for vectorization.            │
+│                                   [default: text-embedding-3-small]                    │
 ╰────────────────────────────────────────────────────────────────────────────────────────╯
 ╭─ Server Configuration ─────────────────────────────────────────────────────────────────╮
 │ --host        TEXT     Host/IP to bind API servers to.                                 │

{agent_cli-0.72.6 → agent_cli-0.73.0}/README.md

@@ -1974,8 +1974,12 @@ uv tool install "agent-cli[vad]" -p 3.13
 │                                [env var: OPENAI_API_KEY]                               │
 ╰────────────────────────────────────────────────────────────────────────────────────────╯
 ╭─ LLM Configuration ────────────────────────────────────────────────────────────────────╮
-│ --embedding-model        TEXT  Embedding model to use for vectorization.               │
-│                                [default: text-embedding-3-small]                       │
+│ --embedding-base-url        TEXT  Base URL for embedding API. Falls back to            │
+│                                   --openai-base-url if not set. Useful when using      │
+│                                   different providers for chat vs embeddings.          │
+│                                   [env var: EMBEDDING_BASE_URL]                        │
+│ --embedding-model           TEXT  Embedding model to use for vectorization.            │
+│                                   [default: text-embedding-3-small]                    │
 ╰────────────────────────────────────────────────────────────────────────────────────────╯
 ╭─ Server Configuration ─────────────────────────────────────────────────────────────────╮
 │ --host        TEXT     Host/IP to bind API servers to.                                 │
@@ -2137,8 +2141,12 @@ The `memory proxy` command is the core feature—a middleware server that gives
 │                                [env var: OPENAI_API_KEY]                               │
 ╰────────────────────────────────────────────────────────────────────────────────────────╯
 ╭─ LLM Configuration ────────────────────────────────────────────────────────────────────╮
-│ --embedding-model        TEXT  Embedding model to use for vectorization.               │
-│                                [default: text-embedding-3-small]                       │
+│ --embedding-base-url        TEXT  Base URL for embedding API. Falls back to            │
+│                                   --openai-base-url if not set. Useful when using      │
+│                                   different providers for chat vs embeddings.          │
+│                                   [env var: EMBEDDING_BASE_URL]                        │
+│ --embedding-model           TEXT  Embedding model to use for vectorization.            │
+│                                   [default: text-embedding-3-small]                    │
 ╰────────────────────────────────────────────────────────────────────────────────────────╯
 ╭─ Server Configuration ─────────────────────────────────────────────────────────────────╮
 │ --host        TEXT     Host/IP to bind API servers to.                                 │

{agent_cli-0.72.6 → agent_cli-0.73.0}/agent_cli/_extras.json

@@ -29,7 +29,8 @@
     "Long-term memory proxy",
     [
       "chromadb",
-      "yaml"
+      "yaml",
+      "pydantic_ai"
     ]
   ],
   "mlx-whisper": [
@@ -43,7 +44,8 @@
   "rag": [
     "RAG proxy (ChromaDB, embeddings)",
     [
-      "chromadb"
+      "chromadb",
+      "pydantic_ai"
     ]
   ],
   "server": [

{agent_cli-0.72.6 → agent_cli-0.73.0}/agent_cli/_requirements/memory.txt

@@ -6,6 +6,7 @@ annotated-types==0.7.0
     # via pydantic
 anyio==4.12.1
     # via
+    #   google-genai
     #   httpx
     #   openai
     #   starlette
@@ -37,16 +38,18 @@ click==8.3.1
     #   typer
     #   typer-slim
     #   uvicorn
-colorama==0.4.6 ; os_name == 'nt' or sys_platform == 'win32'
+colorama==0.4.6
     # via
     #   build
     #   click
+    #   griffe
     #   tqdm
     #   uvicorn
 coloredlogs==15.0.1
     # via onnxruntime
 distro==1.9.0
     # via
+    #   google-genai
     #   openai
     #   posthog
 dnspython==2.8.0
@@ -75,10 +78,18 @@ flatbuffers==25.12.19
     # via onnxruntime
 fsspec==2026.1.0
     # via huggingface-hub
+genai-prices==0.0.51
+    # via pydantic-ai-slim
 google-auth==2.47.0
-    # via kubernetes
+    # via
+    #   google-genai
+    #   kubernetes
+google-genai==1.58.0
+    # via pydantic-ai-slim
 googleapis-common-protos==1.72.0
     # via opentelemetry-exporter-otlp-proto-grpc
+griffe==1.15.0
+    # via pydantic-ai-slim
 grpcio==1.76.0
     # via
     #   chromadb
@@ -99,7 +110,11 @@ httpx==0.28.1
     #   chromadb
     #   fastapi
     #   fastapi-cloud-cli
+    #   genai-prices
+    #   google-genai
     #   openai
+    #   pydantic-ai-slim
+    #   pydantic-graph
 huggingface-hub==0.36.0
     # via
     #   agent-cli
@@ -127,6 +142,8 @@ jsonschema-specifications==2025.9.1
     # via jsonschema
 kubernetes==34.1.0
     # via chromadb
+logfire-api==4.18.0
+    # via pydantic-graph
 markdown-it-py==4.0.0
     # via rich
 markupsafe==3.0.3
@@ -149,13 +166,16 @@ onnxruntime==1.20.1
     #   agent-cli
     #   chromadb
 openai==2.15.0
-    # via agent-cli
+    # via
+    #   agent-cli
+    #   pydantic-ai-slim
 opentelemetry-api==1.39.1
     # via
     #   chromadb
     #   opentelemetry-exporter-otlp-proto-grpc
     #   opentelemetry-sdk
     #   opentelemetry-semantic-conventions
+    #   pydantic-ai-slim
 opentelemetry-exporter-otlp-proto-common==1.39.1
     # via opentelemetry-exporter-otlp-proto-grpc
 opentelemetry-exporter-otlp-proto-grpc==1.39.1
@@ -203,13 +223,21 @@ pydantic==2.12.5
     #   chromadb
     #   fastapi
     #   fastapi-cloud-cli
+    #   genai-prices
+    #   google-genai
     #   openai
+    #   pydantic-ai-slim
     #   pydantic-extra-types
+    #   pydantic-graph
     #   pydantic-settings
+pydantic-ai-slim==1.42.0
+    # via agent-cli
 pydantic-core==2.41.5
     # via pydantic
 pydantic-extra-types==2.11.0
     # via fastapi
+pydantic-graph==1.42.0
+    # via pydantic-ai-slim
 pydantic-settings==2.12.0
     # via fastapi
 pygments==2.19.2
@@ -246,13 +274,18 @@ referencing==0.37.0
     #   jsonschema
     #   jsonschema-specifications
 regex==2026.1.15
-    # via transformers
+    # via
+    #   tiktoken
+    #   transformers
 requests==2.32.5
     # via
+    #   google-auth
+    #   google-genai
     #   huggingface-hub
     #   kubernetes
     #   posthog
     #   requests-oauthlib
+    #   tiktoken
     #   transformers
 requests-oauthlib==2.0.0
     # via kubernetes
@@ -291,13 +324,19 @@ six==1.17.0
     #   posthog
     #   python-dateutil
 sniffio==1.3.1
-    # via openai
+    # via
+    #   google-genai
+    #   openai
 starlette==0.50.0
     # via fastapi
 sympy==1.14.0
     # via onnxruntime
 tenacity==9.1.2
-    # via chromadb
+    # via
+    #   chromadb
+    #   google-genai
+tiktoken==0.12.0
+    # via pydantic-ai-slim
 tokenizers==0.22.2
     # via
     #   chromadb
@@ -323,6 +362,7 @@ typing-extensions==4.15.0
     #   anyio
     #   chromadb
     #   fastapi
+    #   google-genai
     #   grpcio
     #   huggingface-hub
     #   openai
@@ -342,6 +382,8 @@ typing-extensions==4.15.0
 typing-inspection==0.4.2
     # via
     #   pydantic
+    #   pydantic-ai-slim
+    #   pydantic-graph
     #   pydantic-settings
 urllib3==2.3.0
     # via
@@ -363,6 +405,8 @@ watchfiles==1.1.1
 websocket-client==1.9.0
     # via kubernetes
 websockets==15.0.1
-    # via uvicorn
+    # via
+    #   google-genai
+    #   uvicorn
 zipp==3.23.0
     # via importlib-metadata

{agent_cli-0.72.6 → agent_cli-0.73.0}/agent_cli/_requirements/rag.txt

@@ -6,6 +6,7 @@ annotated-types==0.7.0
     # via pydantic
 anyio==4.12.1
     # via
+    #   google-genai
     #   httpx
     #   openai
     #   starlette
@@ -49,10 +50,11 @@ click==8.3.1
     #   uvicorn
 cobble==0.1.4
     # via mammoth
-colorama==0.4.6 ; os_name == 'nt' or sys_platform == 'win32'
+colorama==0.4.6
     # via
     #   build
     #   click
+    #   griffe
     #   tqdm
     #   uvicorn
 coloredlogs==15.0.1
@@ -63,6 +65,7 @@ defusedxml==0.7.1
     # via markitdown
 distro==1.9.0
     # via
+    #   google-genai
     #   openai
     #   posthog
 dnspython==2.8.0
@@ -91,10 +94,18 @@ flatbuffers==25.12.19
     # via onnxruntime
 fsspec==2026.1.0
     # via huggingface-hub
+genai-prices==0.0.51
+    # via pydantic-ai-slim
 google-auth==2.47.0
-    # via kubernetes
+    # via
+    #   google-genai
+    #   kubernetes
+google-genai==1.58.0
+    # via pydantic-ai-slim
 googleapis-common-protos==1.72.0
     # via opentelemetry-exporter-otlp-proto-grpc
+griffe==1.15.0
+    # via pydantic-ai-slim
 grpcio==1.76.0
     # via
     #   chromadb
@@ -115,7 +126,11 @@ httpx==0.28.1
     #   chromadb
     #   fastapi
     #   fastapi-cloud-cli
+    #   genai-prices
+    #   google-genai
     #   openai
+    #   pydantic-ai-slim
+    #   pydantic-graph
 huggingface-hub==0.36.0
     # via
     #   agent-cli
@@ -143,6 +158,8 @@ jsonschema-specifications==2025.9.1
     # via jsonschema
 kubernetes==34.1.0
     # via chromadb
+logfire-api==4.18.0
+    # via pydantic-graph
 lxml==6.0.2
     # via
     #   markitdown
@@ -180,13 +197,16 @@ onnxruntime==1.20.1
     #   magika
     #   markitdown
 openai==2.15.0
-    # via agent-cli
+    # via
+    #   agent-cli
+    #   pydantic-ai-slim
 opentelemetry-api==1.39.1
     # via
     #   chromadb
     #   opentelemetry-exporter-otlp-proto-grpc
     #   opentelemetry-sdk
     #   opentelemetry-semantic-conventions
+    #   pydantic-ai-slim
 opentelemetry-exporter-otlp-proto-common==1.39.1
     # via opentelemetry-exporter-otlp-proto-grpc
 opentelemetry-exporter-otlp-proto-grpc==1.39.1
@@ -240,13 +260,21 @@ pydantic==2.12.5
     #   chromadb
     #   fastapi
     #   fastapi-cloud-cli
+    #   genai-prices
+    #   google-genai
     #   openai
+    #   pydantic-ai-slim
     #   pydantic-extra-types
+    #   pydantic-graph
     #   pydantic-settings
+pydantic-ai-slim==1.42.0
+    # via agent-cli
 pydantic-core==2.41.5
     # via pydantic
 pydantic-extra-types==2.11.0
     # via fastapi
+pydantic-graph==1.42.0
+    # via pydantic-ai-slim
 pydantic-settings==2.12.0
     # via fastapi
 pygments==2.19.2
@@ -285,14 +313,19 @@ referencing==0.37.0
     #   jsonschema
     #   jsonschema-specifications
 regex==2026.1.15
-    # via transformers
+    # via
+    #   tiktoken
+    #   transformers
 requests==2.32.5
     # via
+    #   google-auth
+    #   google-genai
     #   huggingface-hub
     #   kubernetes
     #   markitdown
     #   posthog
     #   requests-oauthlib
+    #   tiktoken
     #   transformers
 requests-oauthlib==2.0.0
     # via kubernetes
@@ -332,7 +365,9 @@ six==1.17.0
     #   posthog
     #   python-dateutil
 sniffio==1.3.1
-    # via openai
+    # via
+    #   google-genai
+    #   openai
 soupsieve==2.8.1
     # via beautifulsoup4
 starlette==0.50.0
@@ -340,7 +375,11 @@ starlette==0.50.0
 sympy==1.14.0
     # via onnxruntime
 tenacity==9.1.2
-    # via chromadb
+    # via
+    #   chromadb
+    #   google-genai
+tiktoken==0.12.0
+    # via pydantic-ai-slim
 tokenizers==0.22.2
     # via
     #   chromadb
@@ -367,6 +406,7 @@ typing-extensions==4.15.0
     #   beautifulsoup4
     #   chromadb
     #   fastapi
+    #   google-genai
     #   grpcio
     #   huggingface-hub
     #   openai
@@ -387,6 +427,8 @@ typing-extensions==4.15.0
 typing-inspection==0.4.2
     # via
     #   pydantic
+    #   pydantic-ai-slim
+    #   pydantic-graph
     #   pydantic-settings
 urllib3==2.3.0
     # via
@@ -408,7 +450,9 @@ watchfiles==1.1.1
 websocket-client==1.9.0
     # via kubernetes
 websockets==15.0.1
-    # via uvicorn
+    # via
+    #   google-genai
+    #   uvicorn
 xlsxwriter==3.2.9
     # via python-pptx
 zipp==3.23.0

{agent_cli-0.72.6 → agent_cli-0.73.0}/agent_cli/agents/memory/proxy.py

@@ -23,6 +23,7 @@ def proxy(
         rich_help_panel="Memory Configuration",
     ),
     openai_base_url: str | None = opts.OPENAI_BASE_URL,
+    embedding_base_url: str | None = opts.EMBEDDING_BASE_URL,
     embedding_model: str = opts.EMBEDDING_MODEL,
     openai_api_key: str | None = opts.OPENAI_API_KEY,
     default_top_k: int = typer.Option(
@@ -135,12 +136,15 @@ def proxy(
     entries_dir, _ = ensure_store_dirs(memory_path)
     if openai_base_url is None:
         openai_base_url = constants.DEFAULT_OPENAI_BASE_URL
+    effective_embedding_url = embedding_base_url or openai_base_url
 
     console.print(f"[bold green]Starting Memory Proxy on {host}:{port}[/bold green]")
     console.print(f"  💾 Memory store: [blue]{memory_path}[/blue]")
     console.print(f"  📁 Entries: [blue]{entries_dir}[/blue]")
     console.print(f"  🤖 Backend: [blue]{openai_base_url}[/blue]")
-    console.print(f"  🧠 Embeddings: Using [blue]{embedding_model}[/blue]")
+    console.print(
+        f"  🧠 Embeddings: [blue]{embedding_model}[/blue] via [blue]{effective_embedding_url}[/blue]",
+    )
     console.print(f"  🔍 Memory top_k: [blue]{default_top_k}[/blue] entries per query")
     console.print(f"  🧹 Max entries per conversation: [blue]{max_entries}[/blue]")
     console.print(
@@ -154,6 +158,7 @@ def proxy(
     fastapi_app = create_app(
         memory_path,
         openai_base_url,
+        embedding_base_url=effective_embedding_url,
         embedding_model=embedding_model,
         embedding_api_key=openai_api_key,
         chat_api_key=openai_api_key,

{agent_cli-0.72.6 → agent_cli-0.73.0}/agent_cli/agents/rag_proxy.py

@@ -32,6 +32,7 @@ def rag_proxy(
         rich_help_panel="RAG Configuration",
     ),
     openai_base_url: str | None = opts.OPENAI_BASE_URL,
+    embedding_base_url: str | None = opts.EMBEDDING_BASE_URL,
     embedding_model: str = opts.EMBEDDING_MODEL,
     openai_api_key: str | None = opts.OPENAI_API_KEY,
     limit: int = typer.Option(
@@ -138,22 +139,26 @@ def rag_proxy(
 
     if openai_base_url is None:
         openai_base_url = constants.DEFAULT_OPENAI_BASE_URL
+    effective_embedding_url = embedding_base_url or openai_base_url
 
     console.print(f"[bold green]Starting RAG Proxy on {host}:{port}[/bold green]")
     console.print(f"  📂 Docs: [blue]{docs_folder}[/blue]")
     console.print(f"  💾 DB: [blue]{chroma_path}[/blue]")
     console.print(f"  🤖 Backend: [blue]{openai_base_url}[/blue]")
-    console.print(f"  🧠 Embeddings: Using [blue]{embedding_model}[/blue]")
+    console.print(
+        f"  🧠 Embeddings: [blue]{embedding_model}[/blue] via [blue]{effective_embedding_url}[/blue]",
+    )
     console.print(f"  🔍 Limit: [blue]{limit}[/blue] chunks per query")
 
     fastapi_app = create_app(
         docs_folder,
         chroma_path,
         openai_base_url,
-        embedding_model,
-        openai_api_key,
-        openai_api_key,
-        limit,
+        embedding_base_url=effective_embedding_url,
+        embedding_model=embedding_model,
+        embedding_api_key=openai_api_key,
+        chat_api_key=openai_api_key,
+        limit=limit,
         enable_rag_tools=enable_rag_tools,
     )
 

{agent_cli-0.72.6 → agent_cli-0.73.0}/agent_cli/memory/api.py

@@ -22,6 +22,7 @@ LOGGER = logging.getLogger(__name__)
 def create_app(
     memory_path: Path,
     openai_base_url: str,
+    embedding_base_url: str | None = None,
     embedding_model: str = DEFAULT_OPENAI_EMBEDDING_MODEL,
     embedding_api_key: str | None = None,
     chat_api_key: str | None = None,
@@ -39,6 +40,7 @@ def create_app(
     client = MemoryClient(
         memory_path=memory_path,
         openai_base_url=openai_base_url,
+        embedding_base_url=embedding_base_url,
         embedding_model=embedding_model,
         embedding_api_key=embedding_api_key,
         chat_api_key=chat_api_key,

{agent_cli-0.72.6 → agent_cli-0.73.0}/agent_cli/memory/client.py

@@ -41,6 +41,7 @@ class MemoryClient:
         self,
         memory_path: Path,
         openai_base_url: str,
+        embedding_base_url: str | None = None,
         embedding_model: str = DEFAULT_OPENAI_EMBEDDING_MODEL,
         embedding_api_key: str | None = None,
         chat_api_key: str | None = None,
@@ -74,7 +75,7 @@ class MemoryClient:
         self.collection: Collection = init_memory_collection(
             self.memory_path,
             embedding_model=embedding_model,
-            openai_base_url=self.openai_base_url,
+            openai_base_url=embedding_base_url or self.openai_base_url,
             openai_api_key=embedding_api_key,
         )
 

{agent_cli-0.72.6 → agent_cli-0.73.0}/agent_cli/opts.py

@@ -109,6 +109,14 @@ EMBEDDING_MODEL: str = typer.Option(
     help="Embedding model to use for vectorization.",
     rich_help_panel="LLM Configuration",
 )
+EMBEDDING_BASE_URL: str | None = typer.Option(
+    None,
+    "--embedding-base-url",
+    envvar="EMBEDDING_BASE_URL",
+    help="Base URL for embedding API. Falls back to `--openai-base-url` if not set. "
+    "Useful when using different providers for chat vs embeddings.",
+    rich_help_panel="LLM Configuration",
+)
 
 # --- ASR (Audio) Configuration ---
 # General ASR

{agent_cli-0.72.6 → agent_cli-0.73.0}/agent_cli/rag/api.py

@@ -32,6 +32,7 @@ def create_app(
     docs_folder: Path,
     chroma_path: Path,
     openai_base_url: str,
+    embedding_base_url: str | None = None,
     embedding_model: str = DEFAULT_OPENAI_EMBEDDING_MODEL,
     embedding_api_key: str | None = None,
     chat_api_key: str | None = None,
@@ -47,7 +48,7 @@ def create_app(
         chroma_path,
         name="docs",
         embedding_model=embedding_model,
-        openai_base_url=openai_base_url,
+        openai_base_url=embedding_base_url or openai_base_url,
         openai_api_key=embedding_api_key,
     )
 

{agent_cli-0.72.6 → agent_cli-0.73.0}/agent_cli/scripts/sync_extras.py

@@ -35,8 +35,8 @@ EXTRA_METADATA: dict[str, tuple[str, list[str]]] = {
     "llm": ("LLM framework (pydantic-ai)", ["pydantic_ai"]),
     "llm-core": ("Minimal LLM support for Docker images", ["pydantic_ai"]),
     # Feature extras
-    "rag": ("RAG proxy (ChromaDB, embeddings)", ["chromadb"]),
-    "memory": ("Long-term memory proxy", ["chromadb", "yaml"]),
+    "rag": ("RAG proxy (ChromaDB, embeddings)", ["chromadb", "pydantic_ai"]),
+    "memory": ("Long-term memory proxy", ["chromadb", "yaml", "pydantic_ai"]),
     "vad": ("Voice Activity Detection (silero-vad)", ["silero_vad"]),
     "whisper": ("Local Whisper ASR (faster-whisper)", ["faster_whisper"]),
     "whisper-mlx": ("MLX Whisper for Apple Silicon", ["mlx_whisper"]),

{agent_cli-0.72.6 → agent_cli-0.73.0}/docker/docker-compose.yml

@@ -46,6 +46,7 @@
 #   RAG_LIMIT          - Number of document chunks per query (default: 3)
 #   RAG_ENABLE_TOOLS   - Enable read_full_document tool (default: true)
 #   EMBEDDING_MODEL    - Embedding model for RAG/memory (default: text-embedding-3-small)
+#   EMBEDDING_BASE_URL - Separate endpoint for embeddings (falls back to OPENAI_BASE_URL)
 #
 #   MEMORY_PORT        - Memory proxy port (default: 8100)
 #   MEMORY_TOP_K       - Number of memories per query (default: 5)
@@ -262,6 +263,7 @@ services:
       - RAG_LIMIT=${RAG_LIMIT:-3}
       - RAG_ENABLE_TOOLS=${RAG_ENABLE_TOOLS:-true}
       - EMBEDDING_MODEL=${EMBEDDING_MODEL:-text-embedding-3-small}
+      - EMBEDDING_BASE_URL=${EMBEDDING_BASE_URL:-}
       - LOG_LEVEL=${LOG_LEVEL:-info}
       # OpenAI-compatible backend configuration
       - OPENAI_BASE_URL=${OPENAI_BASE_URL:-http://ollama:11434/v1}
@@ -295,6 +297,7 @@ services:
       - MEMORY_SUMMARIZATION=${MEMORY_SUMMARIZATION:-true}
       - MEMORY_GIT_VERSIONING=${MEMORY_GIT_VERSIONING:-true}
       - EMBEDDING_MODEL=${EMBEDDING_MODEL:-text-embedding-3-small}
+      - EMBEDDING_BASE_URL=${EMBEDDING_BASE_URL:-}
       - LOG_LEVEL=${LOG_LEVEL:-info}
       # OpenAI-compatible backend configuration
       - OPENAI_BASE_URL=${OPENAI_BASE_URL:-http://ollama:11434/v1}

{agent_cli-0.72.6 → agent_cli-0.73.0}/docs/commands/memory.md

@@ -98,6 +98,7 @@ agent-cli chat --openai-base-url http://localhost:8100/v1 --llm-provider openai
 
 | Option | Default | Description |
 |--------|---------|-------------|
+| `--embedding-base-url` | - | Base URL for embedding API. Falls back to `--openai-base-url` if not set. Useful when using different providers for chat vs embeddings. |
 | `--embedding-model` | `text-embedding-3-small` | Embedding model to use for vectorization. |
 
 ### Server Configuration

{agent_cli-0.72.6 → agent_cli-0.73.0}/docs/commands/rag-proxy.md

@@ -78,6 +78,7 @@ agent-cli chat --openai-base-url http://localhost:8000/v1 --llm-provider openai
 
 | Option | Default | Description |
 |--------|---------|-------------|
+| `--embedding-base-url` | - | Base URL for embedding API. Falls back to `--openai-base-url` if not set. Useful when using different providers for chat vs embeddings. |
 | `--embedding-model` | `text-embedding-3-small` | Embedding model to use for vectorization. |
 
 ### Server Configuration