feat(inference): Add Latitude as inference provider

[gsalberto]

Summary

This PR adds support for Latitude.sh as an inference provider.

About Latitude:

• Recently acquired by Megaport (ASX: MP1), a global Network-as-a-Service leader
• Operating 10,000+ physical servers and 1,000+ GPUs globally
• Combined with Megaport's platform spanning 1,000+ data centers in 26 countries
• Tier-3 data centers with 99.99% SLA

AI Inference Platform Features:

• OpenAI-compatible API at https://api.lsh.ai
• Tool calling support
• Structured output (JSON mode)
• Vision/multimodal model support
• Streaming responses
• Dedicated GPUs with consistent performance

Changes

• Added packages/inference/src/providers/latitude.ts with LatitudeConversationalTask and LatitudeTextGenerationTask
• Added latitude-sh to INFERENCE_PROVIDERS in types.ts
• Added latitude-sh to PROVIDERS_HUB_ORGS mapping
• Added latitude-sh to HARDCODED_MODEL_INFERENCE_MAPPING in consts.ts
• Added Latitude provider to PROVIDERS in getProviderHelper.ts

Related PRs

• Python Client: feat(inference): Add Latitude as inference provider huggingface_hub#3715
• Documentation: Add Latitude.sh as inference provider hub-docs#2180

[gsalberto]

Hi @hanouticelina @Wauplin - Just wanted to check if you have bandwidth to review this PR.
We'd love to add Latitude.sh as an inference provider. Let me know if you need any changes

[gsalberto]

Hi @Wauplin @hanouticelina @SBrandeis - friendly bump on this PR. It's been 2 weeks and we'd love to get Latitude added as an inference provider.

We have production infrastructure ready with multiple models (Llama 3.1, Qwen 2.5/3, DeepSeek R1, Gemma 2) running on dedicated GPUs. Our API supports streaming, tool calling, vision, and structured output.

Happy to make any changes needed. Let us know if there's anything blocking the review. Thanks!

[SBrandeis]

Hey there!
Thank you for your interest in becoming an Inference Provider and for the excellent work you've put into this integration!
We really appreciate the effort.

However, we're currently in a consolidation phase focusing on growing usage of Inference Providers via new features and integrations rather than expanding to new partners. This means we've temporarily paused onboarding new providers while we work on these improvements.

We're not able to provide a specific timeline for when we'll resume new provider onboarding, but we'd love to revisit this integration in the future.

Thanks again for your contribution and understanding!