Skip to content

Supported providers

XLAI focuses on a small set of first-party backends with a shared xlai-core contract. Provider matrices in the repo README change as features land; this page summarizes the intent today.

Chat

BackendCrateNotes
OpenAI-compatible HTTPxlai-backend-openaiOpenAI-style HTTP backend for OpenAI and other providers with matching APIs
OpenRouter Responses APIxlai-backend-openrouterDedicated OpenRouter backend for /responses; avoids assuming full OpenAI parity
Google Geminixlai-backend-geminiGemini HTTP API
llama.cpp (local GGUF)xlai-backend-llama-cppNative inference via vendored llama.cpp
transformers.js (browser)xlai-backend-transformersjsWASM-only; delegates generation to a JS adapter

Speech

CapabilityStatus
OpenAI-compatible transcriptionSupported via xlai-backend-openai
OpenAI-compatible TTSSupported via xlai-backend-openai
OpenRouter speech / media APIsNot yet exposed through a dedicated backend
Local Qwen3 TTS (QTS)Native engine in xlai-qts-core; browser path is staged — see QTS

llama.cpp acceleration matrix

The README tracks CPU/GPU EP support (Metal, Vulkan, CUDA roadmap, etc.). See Support of LLM API providers on GitHub for the live checklist.

Released under the Apache License 2.0.