partitech/php-mistral

Connect to MistralAi inference or to Llama.cpp

v0.0.22 2025-04-17 10:34 UTC

Requires

• php: >=8.3
• knplabs/php-json-schema: ^0.1.0
• symfony/http-client: ^6.4 || ^7.0

Requires (Dev)

• phpunit/phpunit: ^9.5

MIT

README

Warning

⚠️ This branch (psr18) introduces PSR-18 (HTTP Client) support.
The implementation is stable, but still evolving. Expect refinements prior to release.

✅ Integration Overview

Backend	Status	Notes
Mistral	✅ Complete	Allmost complete. last step, fine tuning on the way
VLLM	✅ Complete	Fully functional
Hugging Face	✅ Complete	Includes: inference, text-embeddings-inference, text-generation-inference, Datasets manipulation
Ollama	✅ Complete	Fully functional
llama.cpp	✅ Complete	Fully functional
Anthropic	✅ Complete	Fully functional
Grok	✅ Complete	Fully functional
Gemini	⏸ Deferred	Low demand – integration postponed for now

📘 Documentation Status

Full documentation is currently in development.

In the meantime, all major use cases — including request building, client configuration, and response handling — are illustrated via real-world examples in:

📁 examples/Clients

These are canonical examples, directly mapped to each backend’s official API behavior.
You can rely on them today for actual integration and experimentation.

🛠️ Roadmap & Next Steps

• 🧪 PHPUnit-based unit testing
• 📚 Finalize and publish documentation
• 🧭 Gemini (optional, pending interest)

🤝 Contributors Welcome

The core is solid and ready for real-world testing. If you:

• want to test integrations in your own stack,
• find edge cases or inconsistencies,
• or would like to contribute to docs, adapters, or test coverage...

You're more than welcome to jump in 🙌

🚀 Get Started

• 📖 Read the early docs
• 📂 See real-world usage – every backend is covered 🧰