LLM API Pricing Calculator & Model Comparison
The ocean runs on numbers. Stop guessing what your AI agent costs, which model to pick, or what your SOUL.md should say. Calculate it.
API Cost Calculator
Side-by-side model cost
Model Comparison Matrix
Sort & filter all models
OpenClaw Estimator
Sliders β monthly cost
SOUL.md Generator
Download your agent's soul
Current LLM API Prices β March 2026
| Model | Provider | Input $/M tokens | Output $/M tokens | Context |
|---|---|---|---|---|
| DeepSeek V3-0324 | DeepSeek | $0.2 | $0.77 | 64k |
| MiniMax M2 | MiniMax | $0.255 | $1 | 1M |
| Kimi K2 | Moonshot | $0.57 | $2.3 | 128k |
| Claude Haiku 4.5 | Anthropic | $1 | $5 | 200k |
| GPT-4o | OpenAI | $2.5 | $10 | 128k |
| Claude Sonnet 4.6 | Anthropic | $3 | $15 | 200k |
| Claude Opus 4.6 | Anthropic | $5 | $25 | 200k |
Via OpenRouter Β· March 2026 Β· use the interactive tools below to calculate your exact monthly cost
API Cost Calculator
Enter your daily token volume. See what every major model costs per month, side-by-side. Cheapest auto-highlighted.
100k input + 40k output tokens per day = 4.2M tokens/month Β· prices per million tokens
The matrix below is the same data, but navigable. Sort by output cost, filter by provider, spot which model gives you the most context for the money.
Model Comparison Matrix
All March 2026 pricing. Sortable by cost and context window. Filterable by provider and routing tier.
| Model β | Provider | Input $/M β | Output $/M β | Context β | Tier |
|---|---|---|---|---|---|
DeepSeek V3-0324 | DeepSeek | $0.2βΌ | $0.77βΌ | 64k | Fast |
MiniMax M2 | MiniMax | $0.255 | $1 | 1M | Specialist |
Kimi K2 | Moonshot | $0.57 | $2.3 | 128k | Main |
Claude Haiku 4.5 | Anthropic | $1 | $5 | 200k | Fast |
GPT-4o | OpenAI | $2.5 | $10 | 128k | Main |
Claude Sonnet 4.6 | Anthropic | $3 | $15 | 200k | Main |
Claude Opus 4.6 | Anthropic | $5 | $25 | 200k | Deep |
Prices per million tokens via OpenRouter Β· March 2026 Β· βΌ = cheapest in filtered view
Numbers are only useful when they relate to your setup. The estimator below takes your actual usage β message volume, token size, model mix β and tells you exactly what tier you're operating at.
OpenClaw Cost Estimator
Dial in your daily message volume, token size, and model mix. Get your monthly cost and Tier A/B/C label.
Usage
Model Mix (relative weight β auto-normalised to 100%)
1,500 messages/month
$7.59
/month estimated
Tier C β OptimisedWell-tuned routing. Sharky would approve.
Optimised π
Under $15/month. You've got good routing hygiene β or very light usage. Either way, Sharky approves.Cost sorted. Now identity. Every agent worth running has a soul β a persistent identity file that governs how it communicates, what it can do alone, and what it escalates. Fill in the form. Download the file. Drop it in your workspace.
SOUL.md Template Generator
Fill in your agent's name, mandate, traits, and communication channel. Download a production-ready SOUL.md.
Identity
Mandate
Character
Output
SOUL-agent.md
170 lines Β· Markdown format
Drop this file into your agent's workspace directory. Instruct the agent to read it at every session start. Edit the generated file freely β it's a starting point, not a contract.
# Add to your system prompt:
Read SOUL.md before doing anything else.
These files are your identity and operating rules.
Want the full routing config?
The guides have drop-in OpenClaw snippets for all of this.