The Memory Layer for AI Agents - Drop-in memory infrastructure for AI agents and apps. Context that persists. Built for production.
Python agentic-memoryagentic-memory-systemagentsaiai-agentschatgptgenaillmlong-term-memorymemorymemory-managementpythonragstate-management
Doc Updates (#1843)
Prateek Chhikara5b9b65c3958d
diff --git a/docs/components/llms/models/azure_openai.mdx b/docs/components/llms/models/azure_openai.mdxindex 002d4618e..ddb99322b 100644--- a/docs/components/llms/models/azure_openai.mdx+++ b/docs/components/llms/models/azure_openai.mdx@@ -36,6 +36,35 @@ m = Memory.from_config(config)m.add("Likes to play cricket on weekends", user_id="alice", metadata={"category": "hobbies"})```+We also support the new [OpenAI structured-outputs](https://platform.openai.com/docs/guides/structured-outputs/introduction) model.++```python+import os+from mem0 import Memory++os.environ["LLM_AZURE_OPENAI_API_KEY"] = "your-api-key"+os.environ["LLM_AZURE_DEPLOYMENT"] = "your-deployment-name"+os.environ["LLM_AZURE_ENDPOINT"] = "your-api-base-url"+os.environ["LLM_AZURE_API_VERSION"] = "version-to-use"++config = {+ "llm": {+ "provider": "azure_openai_structured",+ "config": {+ "model": "your-deployment-name",+ "temperature": 0.1,+ "max_tokens": 2000,+ "azure_kwargs" : {+ "azure_deployment" : "",+ "api_version" : "",+ "azure_endpoint" : "",+ "api_key" : ""+ }+ }+ }+}+```+## Config-All available parameters for the `azure_openai` config are present in [Master List of All Params in Config](../config).\ No newline at end of file+All available parameters for the `azure_openai` config are present in [Master List of All Params in Config](../config).diff --git a/docs/components/llms/overview.mdx b/docs/components/llms/overview.mdxindex 362ca3dd0..d0fce6a2e 100644--- a/docs/components/llms/overview.mdx+++ b/docs/components/llms/overview.mdx@@ -23,4 +23,28 @@ To view all supported llms, visit the [Supported LLMs](./models).<Card title="Mistral AI" href="/components/llms/models/mistral_ai"></Card><Card title="Google AI" href="/components/llms/models/google_ai"></Card><Card title="AWS bedrock" href="/components/llms/models/aws_bedrock"></Card>-</CardGroup>\ No newline at end of file+</CardGroup>++## Structured vs Unstructured Outputs++Mem0 supports two types of OpenAI LLM formats, each with its own strengths and use cases:++### Structured Outputs++Structured outputs are LLMs that align with OpenAI's structured outputs model:++- **Optimized for:** Returning structured responses (e.g., JSON objects)+- **Benefits:** Precise, easily parseable data+- **Ideal for:** Data extraction, form filling, API responses+- **Learn more:** [OpenAI Structured Outputs Guide](https://platform.openai.com/docs/guides/structured-outputs/introduction)++### Unstructured Outputs++Unstructured outputs correspond to OpenAI's standard, free-form text model:++- **Flexibility:** Returns open-ended, natural language responses+- **Customization:** Use the `response_format` parameter to guide output+- **Trade-off:** Less efficient than structured outputs for specific data needs+- **Best for:** Creative writing, explanations, general conversation++Choose the format that best suits your application's requirements for optimal performance and usability.diff --git a/docs/platform/quickstart.mdx b/docs/platform/quickstart.mdxindex b307e23f4..95d34605a 100644--- a/docs/platform/quickstart.mdx+++ b/docs/platform/quickstart.mdx@@ -55,7 +55,12 @@ messages = [{"role": "user", "content": "Hi, I'm Alex. I'm a vegetarian and I'm allergic to nuts."},{"role": "assistant", "content": "Hello Alex! I've noted that you're a vegetarian and have a nut allergy. I'll keep this in mind for any food-related recommendations or discussions."}]-client.add(messages, user_id="alex")++# The default output_format is v1.0+client.add(messages, user_id="alex", output_format="v1.0")++# To use the latest output_format, set the output_format parameter to "v1.1"+client.add(messages, user_id="alex", output_format="v1.1")``````javascript JavaScript@@ -63,7 +68,7 @@ const messages = [{"role": "user", "content": "Hi, I'm Alex. I'm a vegetarian and I'm allergic to nuts."},{"role": "assistant", "content": "Hello Alex! I've noted that you're a vegetarian and have a nut allergy. I'll keep this in mind for any food-related recommendations or discussions."}];-client.add(messages, { user_id: "alex" })+client.add(messages, { user_id: "alex", output_format: "v1.1" }).then(response => console.log(response)).catch(error => console.error(error));```@@ -77,15 +82,39 @@ curl -X POST "https://api.mem0.ai/v1/memories/" \{"role": "user", "content": "Hi, I'm Alex. I'm a vegetarian and I'm allergic to nuts."},{"role": "assistant", "content": "Hello Alex! I've noted that you're a vegetarian and have a nut allergy. I'll keep this in mind for any food-related recommendations or discussions."}],- "user_id": "alex"+ "user_id": "alex",+ "output_format": "v1.1"}'```-```json Output+```json Output (v1.0){'message': 'ok'}```+```json Output (v1.1)+{+ "results": [+ {+ "memory": "Name is Alex",+ "event": "ADD"+ },+ {+ "memory": "Is a vegetarian",+ "event": "ADD"+ },+ {+ "memory": "Is allergic to nuts",+ "event": "ADD"+ }+ ]+}+```+</CodeGroup>++<Note> The `add` method supports two output formats: `v1.0` (default) and `v1.1`. To use the latest format, which provides more detailed information about each memory operation, set the `output_format` parameter to `v1.1`: </Note>++#### Short-term memory for a user sessionThese memory instances persist only for the duration of a user session. Ideal for non-repetitive interactions and managing context windows efficiently.@@ -99,7 +128,12 @@ messages = [{"role": "user", "content": "Yes, please! Especially in Tokyo."},{"role": "assistant", "content": "Great! I'll remember that you're interested in vegetarian restaurants in Tokyo for your upcoming trip. I'll prepare a list for you in our next interaction."}]-client.add(messages, user_id="alex123", session_id="trip-planning-2024")++# The default output_format is v1.0+client.add(messages, user_id="alex123", session_id="trip-planning-2024", output_format="v1.0")++# To use the latest output_format, set the output_format parameter to "v1.1"+client.add(messages, user_id="alex123", session_id="trip-planning-2024", output_format="v1.1")``````javascript JavaScript@@ -109,7 +143,7 @@ const messages = [{"role": "user", "content": "Yes, please! Especially in Tokyo."},{"role": "assistant", "content": "Great! I'll remember that you're interested in vegetarian restaurants in Tokyo for your upcoming trip. I'll prepare a list for you in our next interaction."}];-client.add(messages, { user_id: "alex123", session_id: "trip-planning-2024" })+client.add(messages, { user_id: "alex123", session_id: "trip-planning-2024", output_format: "v1.1" }).then(response => console.log(response)).catch(error => console.error(error));```@@ -126,15 +160,35 @@ curl -X POST "https://api.mem0.ai/v1/memories/" \{"role": "assistant", "content": "Great! I'll remember that you're interested in vegetarian restaurants in Tokyo for your upcoming trip. I'll prepare a list for you in our next interaction."}],"user_id": "alex123",- "session_id": "trip-planning-2024"+ "session_id": "trip-planning-2024",+ "output_format": "v1.1"}'```-```json Output+```json Output (v1.0){'message': 'ok'}```++```json Output (v1.1)+{+ "results": [+ {+ "memory": "Planning a trip to Japan next month",+ "event": "ADD"+ },+ {+ "memory": "Interested in vegetarian restaurants in Tokyo",+ "event": "ADD"+ }+ ]+}+```+</CodeGroup>+++#### Long-term memory for agentsAdd a memory layer for the assistants and agents so that their responses remain consistent across sessions.@@ -145,7 +199,12 @@ messages = [{"role": "system", "content": "You are an AI tutor with a personality. Give yourself a name for the user."},{"role": "assistant", "content": "Understood. I'm an AI tutor with a personality. My name is Alice."}]-client.add(messages, agent_id="ai-tutor")++# The default output_format is v1.0+client.add(messages, agent_id="ai-tutor", output_format="v1.0")++# To use the latest output_format, set the output_format parameter to "v1.1"+client.add(messages, agent_id="ai-tutor", output_format="v1.1")``````javascript JavaScript@@ -153,7 +212,7 @@ const messages = [{"role": "system", "content": "You are an AI tutor with a personality. Give yourself a name for the user."},{"role": "assistant", "content": "Understood. I'm an AI tutor with a personality. My name is Alice."}];-client.add(messages, { agent_id: "ai-tutor" })+client.add(messages, { agent_id: "ai-tutor", output_format: "v1.1" }).then(response => console.log(response)).catch(error => console.error(error));```@@ -167,15 +226,36 @@ curl -X POST "https://api.mem0.ai/v1/memories/" \{"role": "system", "content": "You are an AI tutor with a personality. Give yourself a name for the user."},{"role": "assistant", "content": "Understood. I'm an AI tutor with a personality. My name is Alice."}],- "agent_id": "ai-tutor"+ "agent_id": "ai-tutor",+ "output_format": "v1.1"}'```-```json Output+```json Output (v1.0){'message': 'ok'}```++```json Output (v1.1)+{+ "results": [+ {+ "memory": "Name is Alex",+ "event": "ADD"+ },+ {+ "memory": "Is a vegetarian",+ "event": "ADD"+ },+ {+ "memory": "Is allergic to nuts",+ "event": "ADD"+ }+ ]+}+```</CodeGroup>+#### Monitor MemoriesYou can monitor memory operations on the platform dashboard:@@ -188,16 +268,23 @@ You can monitor memory operations on the platform dashboard:Pass user messages, interactions, and queries into our search method to retrieve relevant memories.+<Note> The `search` method supports two output formats: `v1.0` (default) and `v1.1`. To use the latest format, which provides more detailed information about each memory operation, set the `output_format` parameter to `v1.1`: </Note>+<CodeGroup>```python Pythonquery = "What should I cook for dinner today?"-client.search(query, user_id="alex")++# The default output_format is v1.0+client.search(query, user_id="alex", output_format="v1.0")++# To use the latest output_format, set the output_format parameter to "v1.1"+client.search(query, user_id="alex", output_format="v1.1")``````javascript JavaScriptconst query = "What should I cook for dinner today?";-client.search(query, { user_id: "alex" })+client.search(query, { user_id: "alex", output_format: "v1.1" }).then(results => console.log(results)).catch(error => console.error(error));```@@ -208,11 +295,12 @@ curl -X POST "https://api.mem0.ai/v1/memories/search/" \-H "Content-Type: application/json" \-d '{"query": "What should I cook for dinner today?",- "user_id": "alex"+ "user_id": "alex",+ "output_format": "v1.1"}'```-```json Output+```json Output (v1.0)[{"id": "7f165f7e-b411-4afe-b7e5-35789b72c4a5",@@ -235,8 +323,35 @@ curl -X POST "https://api.mem0.ai/v1/memories/search/" \}]```++```json Output (v1.1)+{+ "results": [+ {+ "id": "7f165f7e-b411-4afe-b7e5-35789b72c4a5",+ "memory": "Vegetarian. Allergic to nuts.",+ "input": [+ {+ "role": "user",+ "content": "Hi, I'm Alex. I'm a vegetarian and I'm allergic to nuts."+ },+ {+ "role": "assistant",+ "content": "Hello Alex! I've noted that you're a vegetarian and have a nut allergy. I'll keep this in mind for any food-related recommendations or discussions."+ }+ ],+ "user_id": "alex",+ "hash": "9ee7e1455e84d1dab700ed8749aed75a",+ "metadata": null,+ "created_at": "2024-07-20T01:30:36.275141-07:00",+ "updated_at": "2024-07-20T01:30:36.275172-07:00"+ }+ ]+}+```</CodeGroup>+#### Search using custom filtersOur advanced search allows you to set custom search filters. You can filter by user_id, agent_id, app_id, date, and more.@@ -466,26 +581,33 @@ curl -X GET "https://api.mem0.ai/v1/entities/" \Fetch all memories for a user, agent, or session using the getAll() method.+<Note> The `get_all` method supports two output formats: `v1.0` (default) and `v1.1`. To use the latest format, which provides more detailed information about each memory operation, set the `output_format` parameter to `v1.1`: </Note>++#### Get all memories of a user<CodeGroup>```python Python-user_memories = client.get_all(user_id="alex")+# The default output_format is v1.0+user_memories = client.get_all(user_id="alex", output_format="v1.0")++# To use the latest output_format (v1.1), set the output_format parameter to "v1.1"+user_memories = client.get_all(user_id="alex", output_format="v1.1")``````javascript JavaScript-client.getAll({ user_id: "alex" })+client.getAll({ user_id: "alex", output_format: "v1.1" }).then(memories => console.log(memories)).catch(error => console.error(error));``````bash cURL-curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex" \+curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex&output_format=v1.1" \-H "Authorization: Token your-api-key"```-```json Output+```json Output (v1.0)[{"id":"f38b689d-6b24-45b7-bced-17fbb4d8bac7",@@ -507,6 +629,32 @@ curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex" \}]```++```json Output (v1.1)+{+ "results": [+ {+ "id": "f38b689d-6b24-45b7-bced-17fbb4d8bac7",+ "memory": "是素食主义者,对坚果过敏。",+ "agent_id": "travel-assistant",+ "hash": "62bc074f56d1f909f1b4c2b639f56f6a",+ "metadata": "None",+ "created_at": "2024-07-25T23:57:00.108347-07:00",+ "updated_at": "2024-07-25T23:57:00.108367-07:00"+ },+ {+ "id": "0a14d8f0-e364-4f5c-b305-10da1f0d0878",+ "memory": "Will maintain personalized travel preferences for each user. Provide customized recommendations based on dietary restrictions, interests, and past interactions.",+ "agent_id": "travel-assistant",+ "hash": "35a305373d639b0bffc6c2a3e2eb4244",+ "metadata": "None",+ "created_at": "2024-07-26T00:31:03.543759-07:00",+ "updated_at": "2024-07-26T00:31:03.543778-07:00"+ }+ ]+}+```+</CodeGroup>@@ -514,21 +662,25 @@ curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex" \<CodeGroup>```python Python-client.get_all(agent_id="ai-tutor")+# The default output_format is v1.0+client.get_all(agent_id="ai-tutor", output_format="v1.0")++# To use the latest output_format (v1.1), set the output_format parameter to "v1.1"+client.get_all(agent_id="ai-tutor", output_format="v1.1")``````javascript JavaScript-client.getAll({ agent_id: "ai-tutor" })+client.getAll({ agent_id: "ai-tutor", output_format: "v1.1" }).then(memories => console.log(memories)).catch(error => console.error(error));``````bash cURL-curl -X GET "https://api.mem0.ai/v1/memories/?agent_id=travel-assistant" \+curl -X GET "https://api.mem0.ai/v1/memories/?agent_id=travel-assistant&output_format=v1.1" \-H "Authorization: Token your-api-key"```-```json Output+```json Output (v1.0)[{"id":"f38b689d-6b24-45b7-bced-17fbb4d8bac7",@@ -550,6 +702,31 @@ curl -X GET "https://api.mem0.ai/v1/memories/?agent_id=travel-assistant" \}]```++```json Output (v1.1)+{+ "results": [+ {+ "id": "f38b689d-6b24-45b7-bced-17fbb4d8bac7",+ "memory": "是素食主义者,对坚果过敏。",+ "agent_id": "ai-tutor",+ "hash": "62bc074f56d1f909f1b4c2b639f56f6a",+ "metadata": "None",+ "created_at": "2024-07-25T23:57:00.108347-07:00",+ "updated_at": "2024-07-25T23:57:00.108367-07:00"+ },+ {+ "id": "0a14d8f0-e364-4f5c-b305-10da1f0d0878",+ "memory": "My name is Alice.",+ "agent_id": "ai-tutor",+ "hash": "35a305373d639b0bffc6c2a3e2eb4244",+ "metadata": "None",+ "created_at": "2024-07-26T00:31:03.543759-07:00",+ "updated_at": "2024-07-26T00:31:03.543778-07:00"+ }+ ]+}+```</CodeGroup>#### Get the short-term memories for a session@@ -557,21 +734,25 @@ curl -X GET "https://api.mem0.ai/v1/memories/?agent_id=travel-assistant" \<CodeGroup>```python Python-short_term_memories = client.get_all(user_id="alex123", session_id="trip-planning-2024")+# The default output_format is v1.0+short_term_memories = client.get_all(user_id="alex123", session_id="trip-planning-2024", output_format="v1.0")++# To use the latest output_format (v1.1), set the output_format parameter to "v1.1"+short_term_memories = client.get_all(user_id="alex123", session_id="trip-planning-2024", output_format="v1.1")``````javascript JavaScript-client.getAll({ user_id: "alex123", session_id: "trip-planning-2024" })+client.getAll({ user_id: "alex123", session_id: "trip-planning-2024", output_format: "v1.1" }).then(memories => console.log(memories)).catch(error => console.error(error));``````bash cURL-curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&session_id=trip-planning-2024" \+curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&session_id=trip-planning-2024&output_format=v1.1" \-H "Authorization: Token your-api-key"```-```json Output+```json Output (v1.0)[{"id":"06d8df63-7bd2-4fad-9acb-60871bcecee0",@@ -602,8 +783,43 @@ curl -X GET "https://api.mem0.ai/v1/memories/?user_id=alex123&session_id=trip-pl}]```++```json Output (v1.1)+{+ "results": [+ {+ "id": "06d8df63-7bd2-4fad-9acb-60871bcecee0",+ "memory": "Planning a trip to Japan next month. Interested in vegetarian restaurants in Tokyo.",+ "user_id": "alex123",+ "hash": "d2088c936e259f2f5d2d75543d31401c",+ "metadata": "None",+ "created_at": "2024-07-26T00:25:16.566471-07:00",+ "updated_at": "2024-07-26T00:25:16.566492-07:00"+ },+ {+ "id": "b4229775-d860-4ccb-983f-0f628ca112f5",+ "memory": "Planning a trip to Japan next month. Interested in vegetarian restaurants in Tokyo.",+ "user_id": "alex123",+ "hash": "d2088c936e259f2f5d2d75543d31401c",+ "metadata": "None",+ "created_at": "2024-07-26T00:33:20.350542-07:00",+ "updated_at": "2024-07-26T00:33:20.350560-07:00"+ },+ {+ "id": "df1aca24-76cf-4b92-9f58-d03857efcb64",+ "memory": "Planning a trip to Japan next month. Interested in vegetarian restaurants in Tokyo.",+ "user_id": "alex123",+ "hash": "d2088c936e259f2f5d2d75543d31401c",+ "metadata": "None",+ "created_at": "2024-07-26T00:51:09.642275-07:00",+ "updated_at": "2024-07-26T00:51:09.642295-07:00"+ }+ ]+}+```</CodeGroup>+#### Get specific memory<CodeGroup>diff --git a/mem0/llms/ollama.py b/mem0/llms/ollama.pyindex 6750f0dd4..54d8b719f 100644--- a/mem0/llms/ollama.py+++ b/mem0/llms/ollama.py@@ -85,7 +85,8 @@ class OllamaLLM(LLMBase):},}if response_format:- params["format"] = response_format+ params["format"] = "json"+if tools:params["tools"] = tools