Add LiteLLM, Cerebras, and OpenRouter embedding provider support

[bashandbone]

Description

Add support for LiteLLM, Cerebras, and OpenRouter providers to expand embedding model options.

Current State

• File: openai_factory.py:315
• Limited provider support
• Missing popular alternatives

Providers to Add

LiteLLM

• Unified API for 100+ LLM providers
• Simplifies multi-provider support
• Built-in fallbacks and load balancing
• Cost tracking

Benefits:

• Access to many providers through one interface
• Easier provider switching
• Better resilience

Cerebras

• High-performance inference
• Fast embedding generation
• Cost-effective at scale

Benefits:

• Performance improvements
• Additional provider option
• Competitive pricing

OpenRouter (NEW)

• Recently added embedding model support
• OpenAI-compatible API
• Access to multiple models through one endpoint

Benefits:

• Single API for multiple providers
• Simplified integration (OpenAI-compatible)
• Cost optimization options

Implementation Considerations

All three are OpenAI-compatible providers, which should simplify implementation:

• Reuse OpenAI client patterns
• Provider-specific authentication
• API compatibility verification
• Rate limiting differences
• Error handling per provider
• Configuration management

Testing

• Test with each provider's API
• Validate embedding quality
• Performance benchmarking
• Error scenario testing

Source

• File: openai_factory.py:315
• Branch: 003-our-aim-to

[github-actions]

Claude encountered an error —— View job

---

I'll analyze this and get back to you.