Add SLNG model provider

[LukeAM]

Summary

Add SLNG as a model provider, enabling access to 19 voice AI models (11 TTS + 8 STT) across EU, AU, and Asia regions with sub-100ms latency.

What is SLNG?

SLNG is a regional voice AI infrastructure provider that operates TTS/STT models from Deepgram, Cartesia, Fish Audio, Soniox, Sarvam, Speechmatics, and others across regions where major model labs lack infrastructure.

Key Features:

• 19 models (11 TTS + 8 STT)
• Regional deployment (EU, AU, Asia)
• OpenAI-compatible API
• Sub-100ms latency in supported regions

Documentation: https://docs.slng.ai

Changes

Files Added

1. core/llm/llms/Slng.ts (~164 lines)

   • Implements LLM interface
   • Supports streaming and non-streaming completions
   • Lists 19 available models
   • Handles TTS and STT endpoints
   • SSE parsing for streaming

2. docs/customize/model-providers/more/slng.mdx (Documentation)

   • User configuration guide
   • Complete model list with descriptions
   • Example configurations

Implementation Details

API Integration

• Base URL: https://api.slng.ai/v1
• Authentication: Bearer token (SLNG_API_KEY)
• TTS Endpoint: POST /tts/slng/{model}
• STT Endpoint: POST /stt/slng/{model}
• Streaming: Server-Sent Events (SSE)

Available Models

TTS (11 models):

• Deepgram Aura 2, Sarvam Bulbul v3, Cartesia Sonic 3/3.5
• Fish TTS S2.1 Pro, Gradium TTS, Inworld Max 1.5
• Kugel 2, Murf Falcon, Sarvam Saaras, Soniox TTS RT v1

STT (8 models):

• Deepgram Nova 3 (Standard, Medical, Multi)
• Fish Audio ASR, Gradium STT, Reson8 STT v1
• Soniox Speech AI RT v5, Speechmatics Realtime v2

Testing

• Provider loads without errors
• Models list correctly
• Streaming completions work
• Non-streaming completions work
• TTS generation works
• STT transcription works
• Error handling works properly

Documentation

• SLNG API: https://docs.slng.ai
• Models: https://docs.slng.ai/models
• API Reference: https://docs.slng.ai/api-reference

🤖 Generated with Claude Code

[github-actions]

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. You can sign the CLA by just posting a Pull Request Comment same as the below format.

---

I have read the CLA Document and I hereby sign the CLA

---

_{You can retrigger this bot by commenting recheck in this Pull Request. Posted by the CLA Assistant Lite bot.}