Anthropic

Use Claude models via the direct Anthropic API (AnthropicModel) or the Anthropic Bedrock SDK (AnthropicBedrockModel on AWS).

Option 1: Anthropic via AWS Bedrock SDK

Use Claude through AWS with the Anthropic Bedrock SDK — best when you already run on AWS and want IAM-based access.

For standalone invoke / stream usage, see LLM Models — Anthropic via Bedrock SDK.

Install

pip install --extra-index-url https://core-packages.elsai.ai/root/elsai-model/ elsai-model==2.0.0
pip install --extra-index-url https://elsai-agents.elsai.ai/root/ elsai-agents==0.3.0

Setup

1. Enable Claude model access in the Bedrock console.
2. Configure AWS credentials and export the model ID:

export AWS_ACCESS_KEY_ID=...
export AWS_SECRET_ACCESS_KEY=...
export AWS_REGION=us-east-1
export ANTHROPIC_BEDROCK_MODEL_ID=anthropic.claude-3-5-sonnet-20241022-v2:0

Agent — basic

import os
from elsai import Agent
from elsai_model.anthropic_bedrock import AnthropicBedrockModel

model = AnthropicBedrockModel(
    model_id=os.getenv("ANTHROPIC_BEDROCK_MODEL_ID"),
    region_name=os.getenv("AWS_REGION", "us-east-1"),
    max_tokens=256,
    temperature=0.2,
)
agent = Agent(
    model=model,
    system_prompt="You are a concise assistant.",
)
result = agent("Explain the theory of relativity simply.")
print(result.message["content"][0]["text"])

---

Option 2: Anthropic direct API

Call Claude directly with an Anthropic API key — no AWS required.

For standalone invoke / stream usage, see LLM Models — Anthropic (direct API).

Install

pip install --extra-index-url https://core-packages.elsai.ai/root/elsai-model/ elsai-model==2.0.0
pip install --extra-index-url https://elsai-agents.elsai.ai/root/ elsai-agents==0.3.0

Setup

1. Create an Anthropic API key.
2. Export credentials:

export ANTHROPIC_API_KEY=sk-ant-...
export ANTHROPIC_MODEL_NAME=claude-3-5-sonnet-latest

Agent — basic

import os
from elsai import Agent
from elsai_model.anthropic import AnthropicModel

model = AnthropicModel(
    model_id=os.getenv("ANTHROPIC_MODEL_NAME", "claude-3-5-sonnet-latest"),
    max_tokens=256,
    client_args={"api_key": os.environ["ANTHROPIC_API_KEY"]},
    params={"temperature": 0.2},
)
agent = Agent(
    model=model,
    system_prompt="You are a concise assistant.",
)
result = agent("Explain the theory of relativity simply.")
print(result.message["content"][0]["text"])