feat: add infomaniak provider with 9 models

[abonvalle]

Summary

This PR adds support for providers to the Infomaniak integration and 9 AI models:

• Llama 3.3
• Mistral-Small-3.2-24B-Instruct-2506
• Qwen/Qwen3-VL-235B-A22B-Instruct
• Qwen3-Embedding-8B
• granite-3.1-8b
• Gemma-3n-E4B-it
• Kimi-K2.5
• openai/gpt-oss-120b
• Apertus-70B-Instruct-2509

Details

• Added Infomaniak as a new provider with API endpoint, required environment variables, npm package, and documentation link.
• Added full 9 models metadata and capabilities (w/ pricing, I/O tokens).
• Included multimodal support (text, image, video input; text output).
• Enabled reasoning, tool calling, structured output, and temperature support.

Validation

• Ran local validation successfully with opencode