Provider Overview & Feature Matrix

YAICLI supports a wide range of LLM providers, giving you flexibility to choose the model that best fits your needs. This page provides an overview of all supported providers and their capabilities.

Supported Providers

YAICLI currently integrates with the following LLM providers:

Major Cloud Providers

Provider	Description	Default Base URL	API Documentation
OpenAI	GPT models and advanced AI	https://api.openai.com/v1	Docs
Anthropic (Claude)	Advanced reasoning and safety	https://api.anthropic.com	Docs
Google Gemini	Multimodal AI models	https://generativelanguage.googleapis.com/v1beta/openai	Docs
Cohere	Enterprise-focused models	https://api.cohere.com/v2	Docs
Mistral	European AI models	https://api.mistral.ai/v1	Docs

High-Performance Inference

Provider	Description	Default Base URL	API Documentation
Groq	Ultra-fast inference platform	https://api.groq.com/openai/v1	Docs
Cerebras	Hardware-accelerated inference	https://api.cerebras.ai	Docs
NVIDIA NIM	GPU-optimized models	https://integrate.api.nvidia.com/v1	Docs
SambaNova	High-performance AI platform	https://api.sambanova.ai/v1	Docs

Multi-Provider & Aggregation

Provider	Description	Default Base URL	API Documentation
OpenRouter	Unified API for multiple providers	https://openrouter.ai/api/v1	Docs
Together	Open-source model platform	https://api.together.xyz/v1	Docs
HuggingFace	Open-source model hub	-	Docs

Cloud Platform Integration

Provider	Description	Default Base URL	API Documentation
Vertex AI	Google Cloud AI platform	-	Docs

Specialized & Regional

Provider	Description	Default Base URL	API Documentation
DeepSeek	Code and reasoning specialist	https://api.deepseek.com/v1	Docs
XAI (Grok)	Real-time information models	https://api.xai.com/v1	Docs
Yi (01.AI)	Multilingual AI models	https://api.lingyiwanwu.com/v1	Docs
Doubao	ByteDance AI models	https://ark.cn-beijing.volces.com/api/v3	Docs
ChatGLM	Zhipu AI models	https://open.bigmodel.cn/api/paas/v4/	Docs
Moonshot	Chinese AI models	https://api.moonshot.cn/v1	Docs
Minimax	Chinese multimodal models	https://api.minimaxi.com/v1	Docs
ModelScope	Alibaba's model platform	https://api-inference.modelscope.cn/v1/	Docs
Bailian	Alibaba Cloud AI	https://dashscope.aliyuncs.com/compatible-mode/v1	Docs

Local & Self-Hosted

Provider	Description	Default Base URL	API Documentation
Ollama	Local model hosting	http://localhost:11434	Docs

Feature Comparison

Core Features

Provider	Streaming	Function Calling	MCP Support
OpenAI	✅	✅	✅
Anthropic (Claude)	✅	✅	✅
Google Gemini	✅	✅	✅
Cohere	✅	✅	✅
Mistral	✅	✅	✅
Groq	✅	✅	✅
Cerebras	✅	✅	✅
NVIDIA NIM	✅	✅	✅
SambaNova	✅	✅	✅
OpenRouter	✅	✅	✅
Together	✅	✅	✅
HuggingFace	✅	✅	✅
Vertex AI	✅	✅	✅
DeepSeek	✅	✅	✅
XAI (Grok)	✅	✅	✅
Yi (01.AI)	✅	✅	✅
Doubao	✅	✅	✅
ChatGLM	✅	✅	✅
Ollama	✅	✅	✅

Performance Characteristics

Provider	Speed	Cost	Context	Special Features
Groq	🔥🔥🔥	💰💰	Standard	Ultra-fast inference
Cerebras	🔥🔥🔥	💰💰💰	Standard	Hardware acceleration
OpenAI	🔥🔥	💰💰💰	Large	Industry standard
Anthropic	🔥🔥	💰💰💰	Very Large	Safety-focused
Ollama	🔥	Free	Varies	Local, private
DeepSeek	🔥🔥	💰	Large	Code specialist
Together	🔥🔥	💰	Standard	Open source focus

Legend: 🔥 = Speed level, 💰 = Cost level

Optional Dependencies

Some providers require additional dependencies that can be installed using pip:

# Install all dependencies
pip install 'yaicli[all]'

# Install specific provider dependencies
pip install 'yaicli[ollama,cohere,doubao,huggingface,gemini,mistral,anthropic]'

Provider-Specific Documentation

For detailed configuration options for each provider, refer to the provider-specific pages:

Major Cloud Providers

• OpenAI
• Anthropic (Claude)
• Google Gemini
• DeepSeek
• Cohere
• Mistral
• ChatGLM
• Moonshot
• XAI (Grok)
• Yi (01.AI)
• Doubao
• Minimax

High-Performance Inference

• Groq
• Cerebras
• NVIDIA NIM
• SambaNova

Multi-Provider & Aggregation

• OpenRouter
• Together
• HuggingFace
• ModelScope
• Bailian

Cloud Platform Integration

• Vertex AI

Local & Self-Hosted

• Ollama

Using Custom OpenAI-Compatible Endpoints

Many providers offer OpenAI-compatible endpoints. To use them:

PROVIDER=openai
BASE_URL=https://your-custom-endpoint.com/v1
API_KEY=your-api-key
MODEL=model-name

Switching Providers via Environment Variables

You can temporarily switch providers using environment variables:

export YAI_PROVIDER=gemini
export YAI_API_KEY=AI...
export YAI_MODEL=gemini-2.5-flash
ai "What is quantum computing?"

This overrides your config file settings for the current session.