albert-api
Multi-provider LLM client for Albert CLI — part of the Ternary Intelligence Stack.
What this crate provides
A unified async LLM client that routes to any provider without vendor lock-in:
| Provider | Protocol |
|---|---|
| Anthropic (Claude) | Messages API |
| OpenAI (GPT-4o, o1) | Chat Completions |
| Google (Gemini) | GenerativeLanguage API |
| XAI (Grok) | OpenAI-compatible |
| HuggingFace | Inference API |
| Ollama | Local OpenAI-compatible |
| Azure OpenAI | Chat Completions |
| AWS Bedrock | Messages API |
Also handles OAuth token exchange, model discovery (list_remote_models), and streaming via SSE.
Part of the Albert ecosystem
| Crate | Role |
|---|---|
albert-runtime |
Session, MCP, auth, bash |
albert-api |
This crate — LLM client |
albert-commands |
Slash command library |
albert-tools |
Tool execution layer |
albert-compat |
Manifest extraction harness |
albert-cli |
Binary (albert) |