Providers

Provider Adapters

ArtemisKit supports multiple LLM providers through adapters. Choose the one that fits your needs.

Available Providers

LLM Providers

Provider	Package	Best For
OpenAI	@artemiskit/adapter-openai	GPT models, production
Azure OpenAI	@artemiskit/adapter-openai	Enterprise, compliance
OpenAI-Compatible	@artemiskit/adapter-openai	Groq, Together, Ollama, etc.
Anthropic	@artemiskit/adapter-anthropic	Claude models
Vercel AI	@artemiskit/adapter-vercel-ai	Multi-provider apps

Agentic Adapters

For testing LangChain chains/agents and multi-agent systems:

Adapter	Package	Best For
LangChain	@artemiskit/adapter-langchain	Chains, agents, LCEL, RAG
DeepAgents	@artemiskit/adapter-deepagents	Multi-agent systems

See Agentic Adapters for full documentation on testing AI agents.

Quick Setup

OpenAI

export OPENAI_API_KEY="sk-..."

provider: openai
model: gpt-5

Anthropic

export ANTHROPIC_API_KEY="sk-ant-..."

provider: anthropic
model: claude-sonnet-4-5-20241022

Azure OpenAI

export AZURE_OPENAI_API_KEY="..."
export AZURE_OPENAI_ENDPOINT="https://your-resource.openai.azure.com"

provider: azure-openai
model: gpt-5

providers:
  azure-openai:
    apiKey: ${AZURE_OPENAI_API_KEY}
    resourceName: ${AZURE_OPENAI_RESOURCE_NAME}
    deploymentName: ${AZURE_OPENAI_DEPLOYMENT_NAME}

Vercel AI SDK

export OPENAI_API_KEY="sk-..."  # or other provider keys

provider: vercel-ai
model: openai:gpt-4o  # format: provider:model

OpenAI-Compatible (Groq, Together, Ollama, etc.)

export GROQ_API_KEY="gsk_..."  # or other provider key

provider: openai
model: llama-3.3-70b-versatile

providers:
  openai:
    apiKey: ${GROQ_API_KEY}
    baseUrl: https://api.groq.com/openai/v1

See OpenAI-Compatible Providers for Groq, Together AI, Fireworks, OpenRouter, Ollama, LM Studio, and vLLM setup guides.

Feature Comparison

Feature	OpenAI	Azure	Anthropic	Vercel AI
Streaming	Yes	Yes	Yes	Yes
Function Calling	Yes	Yes	Yes	Varies
Vision	Yes	Yes	Yes	Varies
JSON Mode	Yes	Yes	No	Varies
Max Context	128K	128K	200K	Varies

Configuration Precedence

Settings are applied in this order (highest to lowest priority):

1. CLI flags - --provider, --model
2. Scenario file - provider:, model:
3. Config file - artemis.config.yaml
4. Environment variables
5. Defaults

See Also

• Scenario Format
• Storage Backends