feat: Add ModelsLab ChatGenerator for Haystack

[adhikjoshi]

Summary

Add a new ModelsLabChatGenerator component to the Haystack integrations repository, enabling users to use ModelsLab API for chat completions.

Changes

• Added modelslab-haystack package with a ChatGenerator component
• Supports ModelsLab's OpenAI-compatible chat completions endpoint
• Configurable model, generation parameters (temperature, max_tokens, etc.), and timeout
• Follows the same patterns as other integrations (anthropic, cohere, etc.)

Usage

from haystack_integrations.components.generators.modelslab import ModelsLabChatGenerator
from haystack.dataclasses import ChatMessage

generator = ModelsLabChatGenerator(
    model="meta-llama/Meta-Llama-3.1-70B-Instruct",
    generation_kwargs={
        "temperature": 0.7,
        "max_tokens": 500,
    }
)

messages = [
    ChatMessage.from_system("You are a helpful assistant."),
    ChatMessage.from_user("What is Python?"),
]

result = generator.run(messages=messages)

Testing

• All 6 unit tests pass
• Tests cover: initialization, serialization/deserialization, message conversion, and basic run

About ModelsLab

ModelsLab is an AI API platform offering access to various LLMs, image generation, and more at competitive pricing. Visit https://modelslab.com for more info.

Checklist

• Code follows project conventions
• Tests included and passing
• Documentation added
• No hardcoded API keys (uses MODELSLAB_API_KEY env var)

---

Submitted by: @adhikjoshi

[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.

---

Adhik Joshi 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 @adhikjoshi thank you for your initiative and for opening this pull request! We discussed the new integration in the team and we couldn't agree on merging the new integration at this point in time. Instead, I would like to encourage you to publish it yourself. We just updated our repository template, which makes it very easy to publish a PyPI package of the integration yourself. Very little effort: https://github.com/deepset-ai/custom-component