feat: add LiteLLM as AI gateway provider

[RheagalFire]

Summary

• Adds LiteLLM as a new provider, enabling users to route chat through 100+ LLM providers (OpenAI, Anthropic, Azure, Bedrock, Vertex AI, Groq, Ollama, etc.) via a single LiteLLM proxy.
• Follows the existing provider pattern exactly (same structure as Fireworks/Perplexity - OpenAI-compatible client with custom baseURL).

Motivation

LiteLLM acts as a unified AI gateway. Instead of configuring individual provider API keys, users can point ChatChat at a single LiteLLM proxy that handles provider routing across 100+ providers. Useful for self-hosters who already run a centralized LiteLLM proxy, or users who need providers not yet directly supported (e.g., AWS Bedrock, Vertex AI).

Changes

• lib/provider/LiteLLM.ts - new OpenAI-compatible client pointing at LiteLLM proxy (configurable via LITELLM_BASE_URL and LITELLM_API_KEY env vars)
• config/provider/litellm.ts - model definitions with example LiteLLM model IDs (openai/gpt-4o-mini, anthropic/claude-3-5-sonnet, openai/gpt-4o)
• config/provider/index.ts - added LiteLLM to Provider enum, AllModelId/AllModelName unions, and Providers list
• app/api/chat/messages/route.ts - added case Provider.LiteLLM: using OpenAI-compatible streaming (same pattern as Fireworks/Perplexity)

Tests

Live E2E - basic completion via LiteLLM proxy routing to Claude Sonnet 4.6:

=== Test 1: Basic completion ===
Model: claude-sonnet-4-6
Content: OK
Finish reason: stop

Live E2E - streaming via LiteLLM proxy:

=== Test 2: Streaming ===
Streamed content: Hello!

Streaming chunks accumulated correctly through the same OpenAIStream path ChatChat uses for Fireworks/Perplexity.

Error handling - invalid model:

=== Test 3: Invalid model error handling ===
Handled gracefully, no crash

=== All tests passed ===

All 3 tests use the exact same new OpenAI({ baseURL, apiKey }) client that lib/provider/LiteLLM.ts exports.

Risk / Compatibility

• Additive only. No existing providers modified.
• No new dependencies - uses the existing openai package since LiteLLM proxy serves an OpenAI-compatible API.
• Default proxy URL is http://localhost:4000/v1 (standard LiteLLM proxy port).

Example usage

1. Start a LiteLLM proxy:

pip install litellm
litellm --model anthropic/claude-3-5-sonnet --port 4000

2. Set environment variables in .env.local:

LITELLM_BASE_URL=http://localhost:4000/v1
LITELLM_API_KEY=sk-your-litellm-key

3. Select "LiteLLM" as provider in ChatChat settings, then use any LiteLLM model ID:

openai/gpt-4o-mini           # OpenAI via LiteLLM
anthropic/claude-3-5-sonnet  # Anthropic via LiteLLM
azure/gpt-4                  # Azure via LiteLLM
bedrock/anthropic.claude-3   # AWS Bedrock via LiteLLM
vertex_ai/gemini-pro         # Vertex AI via LiteLLM

Users can also point at a remote LiteLLM proxy for centralized credential management:

LITELLM_BASE_URL=https://my-proxy.example.com/v1
LITELLM_API_KEY=sk-team-key

[RheagalFire]

cc @okisdev