ai-model-reference

AI model pricing varies significantly by provider and model tier. This Skill provides per-1M token costs across OpenAI, Anthropic, Google, and DeepSeek, enabling direct price comparison for your API calls and helping you select the most cost-effective model for your use case.

Each AI provider uses specific model name strings for API calls. This Skill maintains accurate, up-to-date model identifiers and their corresponding context windows, so you can specify the correct model string without trial and error.

Model selection depends on your latency, cost, and quality requirements. This Skill categorizes models by performance tier—flagship, fast, and value—and provides context windows and pricing for each, helping you match model capability to your constraints.

Prompt caching reduces API costs by charging lower rates for cached token reads. This Skill includes per-1M token pricing for both standard and cached requests across providers, enabling you to calculate savings for repeated or long-context queries.

Batch processing often offers discounted pricing compared to standard API calls. This Skill provides model names, standard pricing, and context window details needed to compare batch vs. real-time costs and select the best processing mode for your workload.

AI model pricing and available models change frequently as providers release new capabilities and adjust costs. This Skill maintains current data for model selection and cost comparisons, though you should verify pricing with official provider documentation before finalizing large-scale decisions.