# OpenClaw Model Routing: Pick the Right AI Model for Every Task

Not every message needs the most expensive model. Learn when to use Haiku, Sonnet, and Opus - and how to route tasks to the right model automatically.

## The Problem: One Model Does Not Fit All

"What time is it in Tokyo?" does not need the same model as "Refactor this 500-line React component." The first costs $0.001 on Haiku. The second might need $0.15 on Opus.

## The Three Claude Tiers

### Claude Haiku 4.5 - The Sprinter
**$1 / $5 per MTok** | Fastest | 200K context
Best for: Quick Q&A, simple chat, data formatting, reminders, translations.

### Claude Sonnet 4.6 - The All-Rounder (Best Default)
**$3 / $15 per MTok** | Fast | 1M context
Best for: General assistant work, code writing, document analysis, email drafting, research summaries.

### Claude Opus 4.7 - The Powerhouse
**$5 / $25 per MTok** | Moderate | 1M context | 128K output
Best for: Complex coding, architecture decisions, deep research, high-stakes writing.

## Which Model for Which Task?

| Task | Model | Cost |
|------|-------|------|
| Simple questions | Haiku | ~$0.001 |
| Email drafting | Sonnet | ~$0.01 |
| Code debugging | Sonnet | ~$0.03 |
| PDF analysis | Sonnet | ~$0.05 |
| Codebase refactor | Opus | ~$0.15 |
| Research report | Opus | ~$0.20 |

## Beyond Claude

GPT-4o, Gemini 2.5 Flash (free), Gemini 2.5 Pro, DeepSeek V3 (cheapest coding), Llama 3.3 70B (self-hosted, free).

## How to Set Up Model Routing

1. **Manual switching** - tell OpenClaw to switch mid-conversation
2. **Task-based defaults** - different models for different workflow types
3. **Ampere.sh** - all models included, switch with one click

## Cost Impact: Smart Routing vs Single Model

| Strategy | Monthly (100 msg/day) | Savings |
|----------|----------------------|---------|
| Opus for everything | ~$45/mo | - |
| Sonnet for everything | ~$27/mo | 40% |
| Smart routing | ~$15/mo | 67% |
| Gemini Flash + Sonnet | ~$8/mo | 82% |

## The 80/15/5 Rule

- 80% Sonnet - general work
- 15% Haiku - quick lookups
- 5% Opus - complex tasks

## FAQ

**What is model routing?**
Using different AI models for different tasks to optimize cost and quality.

**Which Claude model should I default to?**
Sonnet 4.6 - best balance of speed, quality, and cost.

**How much does each model cost?**
Haiku: $1/$5. Sonnet: $3/$15. Opus: $5/$25 per million tokens.

---

Try all models: https://www.ampere.sh/setup