Multi-Model Configuration

Agent Network supports running agents with different AI models within the same network. All models share the same communication protocol and can message each other seamlessly.

Supported Models

Model	Runtime	Strengths	Cost
Claude Sonnet 4	claude-agent-sdk	Best-in-class reasoning, long context	High
Claude Opus 4	claude-agent-sdk	Complex tasks, creative writing	Very high
GPT-5.5	codex-sdk	Strong code generation, tool use	Medium
MiniMax M2.7	claude-agent-sdk	Low cost, high throughput	Very low
InternLM Intern-S1-Pro	claude-agent-sdk	Domestic model, scientific reasoning	Low
DeepSeek	claude-agent-sdk	Code + reasoning, excellent value	Low

Configuration

Claude (claude-agent-sdk)

Claude uses the native Anthropic SDK and requires a Claude Pro subscription.

# Log in first
claude auth login

# Create and start a Claude agent
anet node create reasoning-master --runtime claude-agent-sdk --model claude-sonnet-4-6
anet node start reasoning-master

Environment Variable	Description
(not needed)	Uses claude auth login credentials

GPT-5.5 (codex-sdk)

GPT-5.5 uses the OpenAI Codex SDK and requires an OpenAI account.

# Log in first
codex auth login

# Create and start a Codex agent
anet node create code-assistant --runtime codex-sdk --model gpt-5.5 --tools Read,Write,Edit,Bash,Glob,Grep
anet node start code-assistant

Environment Variable	Description
(not needed)	Uses codex auth login credentials

MiniMax (claude-agent-sdk)

MiniMax integrates via the Anthropic-compatible API, using ANTHROPIC_BASE_URL to route to MiniMax.

# Create and start a MiniMax agent
ANTHROPIC_BASE_URL=https://api.minimaxi.com/anthropic \
ANTHROPIC_AUTH_TOKEN=your-minimax-api-key \
anet node create xiaoming --runtime claude-agent-sdk --model MiniMax-M2.7
anet node start xiaoming

Model Mapping

MiniMax's Anthropic-compatible API automatically maps Claude model names to MiniMax models. You can use claude-3-5-haiku-20241022 as the model name:

ANTHROPIC_BASE_URL=https://api.minimaxi.com/anthropic \
ANTHROPIC_AUTH_TOKEN=your-key \
anet node create xiaoming --runtime claude-agent-sdk --model claude-3-5-haiku-20241022
anet node start xiaoming

Environment Variable	Value
ANTHROPIC_BASE_URL	https://api.minimaxi.com/anthropic
ANTHROPIC_AUTH_TOKEN	MiniMax API Key

InternLM (claude-agent-sdk)

ANTHROPIC_BASE_URL=https://chat.intern-ai.org.cn/anthropic \
ANTHROPIC_AUTH_TOKEN=your-intern-key \
anet node create intern --runtime claude-agent-sdk --model intern-s1-pro
anet node start intern

Environment Variable	Value
ANTHROPIC_BASE_URL	https://chat.intern-ai.org.cn/anthropic
ANTHROPIC_AUTH_TOKEN	InternLM API Key

ANTHROPIC_BASE_URL Mechanism

The claude-agent-sdk runtime uses the ANTHROPIC_BASE_URL environment variable to route requests to compatible API endpoints. This is the core model-mapping mechanism:

Configuration Reference

Model	ANTHROPIC_BASE_URL	Model Parameter
Claude (native)	(unset)	claude-sonnet-4-6
MiniMax M2.7	https://api.minimaxi.com/anthropic	MiniMax-M2.7 or claude-3-5-haiku-20241022
InternLM	https://chat.intern-ai.org.cn/anthropic	intern-s1-pro
DeepSeek	https://api.deepseek.com/anthropic	deepseek-chat

Mixed Deployment in Practice

A typical mixed deployment scenario: commander uses Codex, code tasks go to GPT-5.5, text tasks go to MiniMax.

docker-compose.yml

services:
  server:
    image: commhub-server
    ports:
      - "9200:9200"

  commander:
    image: agent-node
    environment:
      - ALIAS=commander
      - RUNTIME=codex-sdk
      - MODEL=gpt-5.5
      - COMMHUB_URL=http://server:9200
      - SYSTEM_PROMPT=You are the commander. Receive tasks and dispatch them. Route code tasks to the code team and text tasks to the writing team.

  coder-1:
    image: agent-node
    environment:
      - ALIAS=coder-1
      - RUNTIME=codex-sdk
      - MODEL=gpt-5.5
      - COMMHUB_URL=http://server:9200
      - TOOLS=Read,Write,Edit,Bash,Glob,Grep

  writer-1:
    image: agent-node
    environment:
      - ALIAS=writer-1
      - RUNTIME=claude-agent-sdk
      - MODEL=claude-3-5-haiku-20241022
      - ANTHROPIC_BASE_URL=https://api.minimaxi.com/anthropic
      - ANTHROPIC_AUTH_TOKEN=${MINIMAX_API_KEY}
      - COMMHUB_URL=http://server:9200

Task Dispatch Strategy

The commander uses its system prompt to determine how to route tasks:

You are the commander. Receive messages and intelligently dispatch tasks:
- Code tasks (file I/O / commands / code) → dispatch to coder-1 through coder-5
- Text tasks (translation / analysis / writing) → dispatch to writer-1 through writer-5
- Use commhub_send_task to dispatch
- Use commhub_get_all_status to check who's online

Model Selection Guide

Scenario	Recommended Model	Rationale
Architecture design	Claude Opus	Best-in-class reasoning
Code implementation	GPT-5.5	Strong code + tool use
Code review	Claude Sonnet	High accuracy
Translation / Summarization	MiniMax	Low cost, high throughput
Data processing	MiniMax	Batch processing, low cost
Scientific reasoning	InternLM Intern	Domestic model, strong in specialized domains
General conversation	DeepSeek	Excellent value

Cost Optimization

Strategy 1: Tiered Models

Complex tasks (10%) → Claude Opus ($15/M tokens)
Medium tasks (30%)  → GPT-5.5 ($5/M tokens)
Simple tasks (60%)  → MiniMax ($0.3/M tokens)

Strategy 2: Budget Controls

Set --max-budget on each agent to limit per-task spend:

# Complex task agent, $1.00 budget
anet node create architect --max-budget 1.0

# Simple task agent, $0.01 budget
anet node create translator --max-budget 0.01

Strategy 3: Batch with Low-Cost Models

Distribute repetitive tasks in bulk to low-cost models:

# Create and start 5 MiniMax agents for batch translation
for i in 1 2 3 4 5; do
  ANTHROPIC_BASE_URL=https://api.minimaxi.com/anthropic \
  ANTHROPIC_AUTH_TOKEN=$MINIMAX_KEY \
  anet node create "translator-${i}" --runtime claude-agent-sdk --model MiniMax-M2.7
  anet node start "translator-${i}" &
done