feat: Add n1n API as LLM provider

[ron-42]

Description

This PR adds support for n1n API as a new LLM provider in mem0. n1n is a robust API aggregation platform that provides access to 400+ large language models through a single OpenAI-compatible API endpoint.

Key Features:

• Single API key access to 400+ models (OpenAI, Anthropic, Google, Meta, and more)
• OpenAI-compatible implementation for seamless integration
• Cost-effective pricing (up to 1/10 of official prices for some models)
• Global availability with no VPN required
• Support for multimodal capabilities (text, images, video, audio)

Implementation Details:

• Added N1NConfig configuration class extending BaseLlmConfig
• Implemented N1NLLM provider class extending LLMBase
• Uses existing OpenAI Python SDK with custom base URL (https://n1n.ai/v1)
• Full support for tool calling and structured outputs
• Proper error handling and API key validation

Motivation: This integration was requested by the n1n team to provide mem0 users with:

• Access to a wide variety of models without managing multiple API keys
• Competitive pricing with unified billing
• Easy model switching for different use cases
• High availability through global infrastructure

Fixes #3579

Type of change

• [x] New feature (non-breaking change which adds functionality)
• [x] Documentation update

How Has This Been Tested?

Unit Tests

• [x] Created comprehensive unit test suite (tests/llms/test_n1n.py)
• [x] 11 test cases covering all functionality:
  • Initialization with API key (config and environment variable)
  • Error handling for missing API key
  • Default model configuration
  • Custom base URL support
  • Response generation without tools
  • Response generation with tools
  • Custom response format handling
  • Config conversion from BaseLlmConfig
  • Dict-based configuration
  • Multiple model support
• [x] All tests use mocks to avoid actual API calls
• [x] All 11 tests passing ✅

Test Results:

$ pytest tests/llms/test_n1n.py -v
========================= test session starts =========================
collected 11 items

tests/llms/test_n1n.py::test_n1n_initialization_with_api_key PASSED
tests/llms/test_n1n.py::test_n1n_initialization_with_env_var PASSED
tests/llms/test_n1n.py::test_n1n_initialization_without_api_key PASSED
tests/llms/test_n1n.py::test_n1n_default_model PASSED
tests/llms/test_n1n.py::test_n1n_custom_base_url PASSED
tests/llms/test_n1n.py::test_generate_response_without_tools PASSED
tests/llms/test_n1n.py::test_generate_response_with_tools PASSED
tests/llms/test_n1n.py::test_generate_response_with_response_format PASSED
tests/llms/test_n1n.py::test_n1n_config_conversion_from_base PASSED
tests/llms/test_n1n.py::test_n1n_config_from_dict PASSED
tests/llms/test_n1n.py::test_n1n_multiple_models PASSED

========================= 11 passed in 1.25s =========================

Integration Testing

• [x] Tested with real n1n API key
• [x] Verified successful response generation
• [x] Tested with multiple models (gpt-4o-mini, claude-3-5-sonnet)
• [x] Confirmed environment variable handling
• [x] Validated error messages for missing credentials

Test Configuration

from mem0 import Memory

config = {
    "llm": {
        "provider": "n1n",
        "config": {
            "model": "gpt-4o-mini",
            "api_key": "your-n1n-api-key",
            "temperature": 0.1,
            "max_tokens": 2000,
        }
    }
}

m = Memory.from_config(config)
# Tested: add, search, get_all operations ✅

Checklist:

• [x] My code follows the style guidelines of this project
• [x] I have performed a self-review of my own code
• [x] I have commented my code, particularly in hard-to-understand areas
• [x] I have made corresponding changes to the documentation
• [x] My changes generate no new warnings
• [x] I have added tests that prove my fix is effective or that my feature works
• [x] New and existing unit tests pass locally with my changes
• [x] Any dependent changes have been merged and published in downstream modules
• [x] I have checked my code and corrected any misspellings

Additional Information

Files Changed

New Files (4):

• mem0/configs/llms/n1n.py - Configuration class for n1n
• mem0/llms/n1n.py - N1NLLM provider implementation
• tests/llms/test_n1n.py - Comprehensive unit tests
• docs/components/llms/models/n1n.mdx - User documentation

Modified Files (2):

• mem0/utils/factory.py - Registered n1n in LlmFactory
• mem0/llms/configs.py - Added n1n to supported providers

Dependencies

• No new dependencies required ✅
• Uses existing openai package (already in project dependencies)
• Fully compatible with current mem0 architecture

Available Models

Users can access 400+ models including:

• OpenAI: gpt-4o, gpt-4o-mini, gpt-4-turbo, o1, o3-mini
• Anthropic: claude-3-5-sonnet-20241022, claude-3-5-haiku-20241022
• Google: gemini-2.0-flash-exp, gemini-1.5-pro
• Meta: llama-3.3-70b-instruct, llama-3.1-405b-instruct
• And 390+ more models...

Resources

• Website: https://n1n.ai/
• Documentation: https://docs.n1n.ai/
• Get API Key: https://n1n.ai/console
• Pricing: https://n1n.ai/pricing

Backward Compatibility

• [x] Fully backward compatible
• [x] No breaking changes to existing code
• [x] Follows established LLM provider patterns

---

Maintainer Checklist

• [ ] closes #3579
• [ ] Made sure Checks passed

[parshvadaftari]

Hey @ron-42 Thanks for this PR, but n1n is not on the roadmap for now, but will be added in the future. Really appreciate your efforts.

[ron-42]

So @parshvadaftari, do i have to close the PR?

[parshvadaftari]

No you can keep it open, once we have this on roadmap we'll take this PR into consideration.