LLMs

info

Conceptual Guide

Large Language Models (LLMs) are a core component of LangChain. LangChain does not serve its own LLMs, but rather provides a standard interface for interacting with many different LLMs.

All LLMs implement the llms.Model interface:

// Model is an interface multi-modal models implement.
type Model interface {
	// GenerateContent asks the model to generate content from a sequence of
	// messages. It's the most general interface for multi-modal LLMs that support
	// chat-like interactions.
	GenerateContent(ctx context.Context, messages []MessageContent, options ...CallOption) (*ContentResponse, error)

	// Call is a simplified interface for a text-only Model, generating a single
	// string response from a single string prompt.
	//
	// Deprecated: this method is retained for backwards compatibility. Use the
	// more general [GenerateContent] instead.
	Call(ctx context.Context, prompt string, options ...CallOption) (string, error)
}

The interface provides two methods:

• GenerateContent: The modern, recommended method that supports multi-modal inputs and complex message sequences
• Call: A legacy method for simple text-to-text generation (deprecated but still supported)

For backwards compatibility, llms.LLM is provided as a type alias:

// LLM is an alias for model, for backwards compatibility.
// Deprecated: This alias may be removed in the future; please use Model instead.
type LLM = Model

Supported providers

Each provider lives in its own package under github.com/vxcontrol/langchaingo/llms and implements the same llms.Model interface, so switching providers only changes construction, not your calling code:

Provider	Package	Notes
OpenAI	llms/openai	Also the client for OpenAI-compatible APIs (Groq, DeepSeek, OpenRouter, NVIDIA, Ollama Cloud, …) via WithBaseURL
Anthropic	llms/anthropic	Claude models, extended thinking
Google Gemini	llms/googleai	Google AI Studio endpoint
Vertex AI	llms/googleai/vertex	Gemini through Google Cloud
AWS Bedrock	llms/bedrock	Claude, Llama, Titan, Nova and more
Mistral	llms/mistral
Ollama	llms/ollama	Local models
Hugging Face	llms/huggingface
Fake	llms/fake	Scripted responses for tests

Provider-neutral call options

Options passed to GenerateContent (or llms.GenerateFromSinglePrompt) work across providers; the adapter maps each to the target provider's wire format:

• Sampling: llms.WithTemperature, llms.WithTopP, llms.WithTopK, llms.WithMaxTokens, llms.WithN.
• Reasoning (thinking): llms.WithReasoning, llms.WithAdaptiveReasoning, llms.WithReasoningDisabled — see Configure LLM Providers.
• Structured output: llms.WithStructuredOutput requests schema-constrained JSON and validates the response against your schema.
• Streaming: llms.WithStreamingFunc receives a streaming.Chunk per token.
• Tools / function calling: llms.WithTools.

🗃️ Integrations

10 items