AI SDK Providers
The AI SDK comes with several providers that you can use to interact with different language models:
OpenAI
Image InputObject GenerationTool UsageTool Streaming
Azure
Image InputObject GenerationTool UsageTool Streaming
Anthropic
Image InputObject GenerationTool UsageTool Streaming
Amazon Bedrock
Image InputObject GenerationTool UsageTool Streaming
Google Generative AI
Image InputObject GenerationTool UsageTool Streaming
Google Vertex AI
Image InputObject GenerationTool UsageTool Streaming
Mistral
Image InputObject GenerationTool UsageTool Streaming
Groq
Object GenerationTool UsageTool Streaming
Perplexity
Object GenerationTool UsageTool Streaming
Fireworks
Object GenerationTool UsageTool Streaming
Cohere
There are also community providers that have been created using the Language Model Specification.
LLamaCpp
Provider Dependent
Ollama
Provider Dependent
Chrome AI
Provider Dependent
Anthropic Vertex
Provider Dependent
Portkey
Provider Dependent
Cloudflare Workers AI
Provider Dependent
Write your own
Provider Dependent
Provider support
Not all providers support all AI SDK features. Here's a quick comparison of the capabilities of popular models:
Provider | Model | Image Input | Object Generation | Tool Usage | Tool Streaming |
---|---|---|---|---|---|
OpenAI | gpt-4o | ||||
OpenAI | gpt-4o-mini | ||||
OpenAI | gpt-4-turbo | ||||
OpenAI | gpt-4 | ||||
OpenAI | o1-preview | ||||
OpenAI | o1-mini | ||||
Anthropic | claude-3-5-sonnet-20241022 | ||||
Anthropic | claude-3-5-sonnet-20240620 | ||||
Anthropic | claude-3-5-haiku-20241022 | ||||
Mistral | mistral-large-latest | ||||
Mistral | mistral-small-latest | ||||
Mistral | pixtral-12b-2409 | ||||
Google Generative AI | gemini-1.5-flash | ||||
Google Generative AI | gemini-1.5-pro | ||||
Google Vertex | gemini-1.5-flash | ||||
Google Vertex | gemini-1.5-pro | ||||
Groq | llama-3.1-405b-reasoning | ||||
Groq | llama-3.1-70b-versatile | ||||
Groq | llama-3.1-8b-instant | ||||
Groq | mixtral-8x7b-32768 | ||||
Groq | gemma2-9b-it |
This table is not exhaustive. Additional models can be found in the provider documentation pages and on the provider websites.