Guides to AI costs
Practical guides to choosing models, estimating costs and running models yourself.
- Guide6 Nov 2026
How to Cut Your AI API Bill in Half: 8 Tactics That Work
Prompt caching, batch jobs, model routing, shorter prompts and output, off-peak hours and long-prompt limits: eight ways to cut AI API costs, with numbers.
- Guide6 Nov 2026
How to Get Your First AI API Key: OpenAI, Claude, Gemini and More
How to create an API key for OpenAI, Claude, Gemini, xAI or DeepSeek, which have a free tier, how billing works, and how to keep your key safe.
- Guide6 Nov 2026
AI Token Cost: What One Token Costs in November 2026
What an AI token costs today, per token and per 1M, on every current model: what a message, an email, a report and a book cost, and why prices differ so much.
- Guide13 Oct 2026
Building Your First AI Chatbot: Model Choice, Code and Real Costs
Build your first AI chatbot with an API: which model to start with, a working code example, how conversation history drives cost, and what it really costs.
- Guide13 Oct 2026
How AI Model Pricing Works: Tokens, Input and Output, Caching and Batch
What a token is, why output costs more than input, and how caching, batch discounts, long-prompt surcharges and off-peak rates change what you pay.
- Guide2 Oct 2026
The Cheapest AI Models in 2026, Ranked by Price
The cheapest AI APIs ranked by price per 1M tokens: small, workhorse, frontier and open-weight models. Tables update automatically from verified prices.
- Guide2 Oct 2026
How Much Does Hermes Agent Cost to Run? A Token Pricing Guide
Hermes Agent is free, but the tokens behind it are not. Monthly costs from under $2 to $150 depending on the model you connect, and how to cut the bill.
- Guide2 Oct 2026
Hermes Agent vs OpenClaw: Which Self-Hosted AI Assistant Costs Less to Run?
Hermes Agent vs OpenClaw, the leading self-hosted AI assistants: features, chat apps, sandboxing and what each costs to run per month in API tokens.
- Guide2 Oct 2026
OpenClaw Setup 101: Your Personal AI Agent
Install OpenClaw, connect a model with your own API key, add WhatsApp or Telegram, install skills and keep it secure, plus which models to use.
- Guide2 Oct 2026
10 Powerful Ways to Use OpenClaw
Ten ways people use OpenClaw: inbox triage, code review, research, meeting prep, smart-home control and more, with what each costs to run in API tokens.
- Guide24 Mar 2026
Qwen3.5-9B + Nemotron Nano 4B: Best Settings for OpenClaw
318 benchmarks to find the best settings for Qwen3.5-9B (agent) and Nemotron-3-Nano-4B (router) in OpenClaw via LM Studio, with copy-paste configs.
Latest news
All news- Claude Haiku 5.5 Launch: Cheap Tokens and the 100K-Prompt Catch22 Oct
- Claude Opus 5.5 Launch: 20% Cheaper Than Opus 522 Oct
- Claude Sonnet 5.5 Launch: Benchmarks, Speed and Cost per Task22 Oct
- September 2026 AI Pricing Roundup: GPT-6, Opus 5.5 and Cheaper Flash Models22 Oct
- Anthropic Suspends Fable 5 and Mythos 5 Access After US Government Directive27 Sep