feat: add MiniMaxChatGenerator component

[octo-patch]

Summary

This PR adds MiniMaxChatGenerator, a new Haystack component that integrates MiniMax LLMs via MiniMax's OpenAI-compatible API.

• Add MiniMaxChatGenerator to haystack/components/generators/chat/
• Supports MiniMax-M2.7 (default) and MiniMax-M2.7-highspeed models
• Uses MINIMAX_API_KEY environment variable for authentication
• Default base URL: https://api.minimax.io/v1
• Inherits all capabilities from OpenAIChatGenerator: streaming, tool calling, serialization/deserialization, async support
• Add unit tests covering initialization, serialization, and round-trip deserialization
• Add release note

Usage

from haystack.components.generators.chat import MiniMaxChatGenerator
from haystack.dataclasses import ChatMessage

# Set MINIMAX_API_KEY environment variable
client = MiniMaxChatGenerator()  # uses MiniMax-M2.7 by default
response = client.run([ChatMessage.from_user("Hello!")])
print(response["replies"][0].text)

API Reference

• Chat (OpenAI Compatible): https://platform.minimax.io/docs/api-reference/text-openai-api

[vercel]

Someone is attempting to deploy a commit to the deepset Team on Vercel.

A member of the Team first needs to authorize it.

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.

---

octo-patch seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[julian-risch]

Hi @octo-patch Thank you for opening this pull request. Without the CLA we can't merge this PR. Either way, I recommend to publish and release this integration yourself using our repository template if you think it's beneficial to users. https://github.com/deepset-ai/custom-component
I'd argue users can use the OpenAIChatGenerator class and set the custom base url themselves?
Regarding the implementation, one thing I noticed is that there shouldn't be a need to re-implement to_dict.