Function cache_read_multiplier 

pub fn cache_read_multiplier(model: &str) -> f64

Multiplier applied to input_price for cache-read tokens. Varies by provider family:

• Anthropic / Bedrock Claude — 10% (0.1×)
• OpenAI (GPT-4o, GPT-5, Codex) — 50% (0.5×)
• Google Gemini — 25% (0.25×)
• Z.AI GLM / Moonshot — 20% (0.2×) (best-effort default)

Returns 0.1 as a conservative default so pricing never silently over-counts for an unknown model.