Add Nextbit provider

[deaquino]

This pull request adds support for several new models and providers in the nextbit integration. It introduces configuration files for a variety of large language models, specifying their names, families, release dates, capabilities, cost structure, context limits, and supported modalities. Additionally, a new provider configuration for Nextbit is included.

New Model Additions:

• Added configuration files for multiple new models, including ALIA Salamandra 40B, Llama 3.3 Euryale 70B, Gemma 2 27B IT, Ministral 3 14B 2512, Ministral 3 8B 2512, MythoMax 13B, Qwen3 30B-A3B, Qwen3.5 9B, ReMM SLERP L2 13B, Rocinante 12B, and UnslopNemo 12B, each specifying model metadata, capabilities, cost, limits, and modalities.

• Introduced several model configuration files that extend from existing base models, customizing cost and context/output limits for models such as deepseek:qwen3-r1-32b, gemma4:26b-a4b, gpt-oss:20b, llama3.3:70b, ministral:3b, qwen3:14b, and qwen:3.5-35b.

Provider Configuration:

• Added a new provider configuration file provider.toml for Nextbit, defining the provider's name, required environment variables, npm package, API endpoint, and documentation URL.

Microsoft Model Support:

• Added support for the Microsoft phi-4 model, including cost and context/output limits.