Releases: lfnovo/esperanto
Support Azure Open AI embeddings
Support AZURE_OPENAI_API_VERSION as an alternative
v2.0.2 fix: accept AZURE_OPENAI_API_VERSION as an alternative env variable
Support OLLAMA_API_BASE as an env variable
To maintain compatibility with existing Ollama installations
v2.0.0 - Major Release: HTTPX Migration
🚀 Esperanto v2.0.0 - Major Release
🔥 Breaking Changes & Major Improvements
This is a major release that fundamentally transforms Esperanto's architecture for production-grade performance and reliability.
⚡ HTTPX Migration - Lightweight & Fast
- Replaced all vendor SDKs with direct HTTPX calls across all 15+ providers
- Ultra-lightweight footprint - no more bulky vendor dependencies
- Production-ready performance - direct HTTP calls mean faster response times
- Consistent error handling - standardized HTTP error management across all providers
🪶 Minimal Dependencies
- Removed SDK bloat - OpenAI, Anthropic, Google, Groq, Mistral, ElevenLabs, and other vendor SDKs eliminated
- Core dependencies only:
httpx,pydantic- that's it! - Optional extras preserved: Only
transformersandlangchainremain as extras
🔄 True Provider Flexibility
- Zero code changes required - same API, same response objects, same methods
- Switch providers instantly - no SDK conflicts or dependency hell
- Future-proof architecture - add new providers without SDK dependencies
📚 Enhanced Documentation
- Comprehensive provider guides - LLM, Embedding, Speech-to-Text, Text-to-Speech
- Google Multi-Speaker TTS - documented unique conversation features
- Production examples - real-world usage patterns and best practices
- Unified selling proposition - clear competitive advantages
🧪 Robust Testing
- Complete test suite migration - all mocking updated for HTTPX patterns
- Credential isolation - tests run without requiring real API keys
- CI/CD ready - reliable testing in all environments
📋 Migration Guide
✅ No Code Changes Required
Your existing code continues to work exactly as before:
from esperanto.factory import AIFactory
# This still works exactly the same\!
model = AIFactory.create_language("openai", "gpt-4")
response = model.chat_complete(messages)📦 Installation Simplified
# Core installation (much lighter now\!)
pip install esperanto
# Optional transformers support
pip install "esperanto[transformers]"🔧 LangChain Integration
Now requires manual installation of LangChain packages:
pip install "langchain>=0.3.8,<0.4.0" "langchain-core>=0.3.29,<0.4.0"
# Then install only the provider packages you need🎯 Performance Benefits
- Faster startup - no SDK initialization overhead
- Lower memory usage - eliminated vendor SDK bloat
- Better error handling - consistent HTTP error patterns
- Easier debugging - direct HTTP requests, no SDK abstraction layers
🛠 What's Changed
- Migrated all providers from vendor SDKs to direct HTTPX HTTP calls
- Removed README.pypi.md duplication (now uses single README.md)
- Updated all documentation with standardized provider patterns
- Enhanced selling proposition highlighting production-ready architecture
- Fixed all test mocking for HTTPX patterns
- Cleaned up dependencies and optional extras
🙏 Acknowledgments
This release represents a complete architectural overhaul focused on production performance and developer experience. The migration from vendor SDKs to HTTPX provides the performance of direct API calls with the convenience of a unified interface.
From prototype to production - same code, better performance.
v1.5.0
What's New
- Mistral Integration: Added support for Mistral as both a Language Model provider (Mistral Large, Small, etc.) and Embedding provider (mistral-embed, etc.). Includes support for chat completions, streaming, JSON mode, and text embeddings. LangChain integration is also available.
- DeepSeek Integration: Added DeepSeek as a Language Model provider. Supports chat completions, streaming, JSON mode, and LangChain integration. Uses the OpenAI-compatible API and can be accessed via the unified Esperanto interface.
See the documentation for usage details and configuration instructions for both new providers.
Other improvements and fixes may be included in this release.
Esperanto v1.4
- Added support for Azure OpenAI
- Fixed Google GenAI provider
- Fixed Google Cloud TTS voices issue
v1.2.1
Full Changelog: v1.2.0...v1.2.1
v1.2.0
Enabled support for Perplexity
Full Changelog: v1.1.2...v1.2.0
Release v1.1.2: Update Google SDK implementation
This release updates the Google SDK implementation to use google-genai v1.8.0, replacing the deprecated google-generativeai package.