Models API

Agent-grade model providers live in elsai-model (elsai Core). Pass a *Model instance — or one built with the LLM factory — to Agent(model=...). The agent loop calls the model on every turn, handling tool specs, system prompts, and streaming events uniformly across providers.

For install, credentials, and full copy-paste kits, see Model Providers. For standalone invoke / stream without agents, see LLM Models.

Signature

import os
from elsai import Agent
from elsai_model import LLM, Provider
# or: from elsai_model.openai import OpenAIModel

agent = Agent(
    model=LLM(
        provider=Provider.OPENAI,
        model="gpt-4o",
        api_key=os.environ["OPENAI_API_KEY"],
    )
)
# agent = Agent(model=OpenAIModel(model_id="gpt-4o", client_args={"api_key": ...}))

Install with a provider extra (example: "elsai-model[openai]==2.1.0"). See LLM Models — Optional package extras.

Supported providers

Provider	Model class	Import module	Factory
OpenAI	OpenAIModel	elsai_model.openai	Provider.OPENAI
Azure OpenAI	AzureOpenAIModel	elsai_model.azure_openai	Provider.AZURE_OPENAI
Amazon Bedrock	BedrockModel	elsai_model.bedrock	Provider.BEDROCK
Anthropic (direct)	AnthropicModel	elsai_model.anthropic	Provider.ANTHROPIC
Anthropic via Bedrock SDK	AnthropicBedrockModel	elsai_model.anthropic_bedrock	Provider.ANTHROPIC_BEDROCK
Google Gemini	GeminiModel	elsai_model.gemini	Provider.GEMINI
LiteLLM	LiteLLMModel	elsai_model.litellm	Provider.LITELLM
Ollama	OllamaModel	elsai_model.ollama	Provider.OLLAMA
Mistral	MistralModel	elsai_model.mistral	Provider.MISTRAL
Writer	WriterModel	elsai_model.writer	Provider.WRITER
Meta Llama API	LlamaAPIModel	elsai_model.llamaapi	Provider.LLAMA_API
llama.cpp server	LlamaCppModel	elsai_model.llama_cpp	Provider.LLAMA_CPP
OpenAI Responses	OpenAIResponsesModel	elsai_model.openai_responses	Provider.OPENAI_RESPONSES
SageMaker	SageMakerAIModel	elsai_model.sagemaker	Provider.SAGEMAKER

Per-provider constructor parameters, environment variables, and examples are documented in LLM Models and the Model Providers concept pages.

Common parameters

Parameter	Description
provider	Provider enum or string when using LLM(provider=…, model=…)
model_id	Provider model or deployment ID (on *Model classes)
model	Same role when using LLM(provider=…, model=…)
api_key	Factory shorthand merged into client_args
client_args	SDK client options (API keys, endpoints, regions)
params	Generation settings (temperature, max_tokens, …)

Factory examples (signatures)

import os
from elsai_model import LLM, Provider

LLM(
    provider=Provider.BEDROCK,
    model="global.anthropic.claude-sonnet-4-6",
    region_name="us-west-2",
)
LLM(
    provider=Provider.OPENAI,
    model="gpt-4o",
    api_key=os.environ["OPENAI_API_KEY"],
    params={"temperature": 0.7},
)

See Amazon Bedrock and OpenAI for full kits.

Legacy connectors

Legacy v1.4.x *Connector classes and GeminiService are not supported with Agent. Pass agent-grade *Model classes from elsai-model==2.1.0 (or build them with LLM) instead. See elsai Core v3.0 release notes.

Related

• LLM Models — LLM factory
• Model Providers — install, credentials, and provider selection
• LLM Models — full provider API reference for elsai-model
• Agent API — Agent(model=…) and runtime options