Prompt caching savings calculator
Pick a model, set your cache hit rate, and see the monthly difference.
Data verified 2026-10-13 from official pricing pages.
Calculator
Without caching
—
With caching
—
You save
—
Cached vs uncached input prices
| Model | Vendor | Input /1M | Cached /1M | Output /1M | Notes | Verified |
|---|---|---|---|---|---|---|
| Claude Opus 5.5 | Anthropic | $4 | $0.2 | $20 | Cache-read figure parsed from page layout; confirm | source · 2026-10-13 verified |
| Claude Opus 5 (legacy) | Anthropic | $5 | $0.5 | $25 | source · 2026-10-13 verified | |
| Claude Sonnet 4.6 (legacy) | Anthropic | $3 | $0.3 | $15 | source · 2026-10-13 verified | |
| GPT-6 Astra | OpenAI | $10 | $1 | $50 | ≤272K context | source · 2026-10-13 verified |
| GPT-6.1 Sol | OpenAI | $2 | $0.1 | $10 | ≤272K context | source · 2026-10-13 verified |
| GPT-6 Luna | OpenAI | $0.1 | $0.01 | $0.5 | source · 2026-10-13 verified | |
| GPT-5.6 Sol | OpenAI | $4 | $0.4 | $20 | Promotional pricing at least through Nov 21, 2026 | source · 2026-10-13 verified |
| GPT-5.3 Codex | OpenAI | $1.75 | $0.175 | $14 | source · 2026-10-13 verified | |
| Gemini 4 Argon | $2 | $0.1 | $10 | Doubles to $4/$20 on Dec 1, 2026 | source · 2026-10-13 verified | |
| Gemini 3.8 Flash | $0.75 | $0.075 | $3.75 | Doubles to $1.50/$7.50 on Jan 1, 2027 | source · 2026-10-13 verified | |
| Gemini 3.1 Pro Preview | $2 | $0.2 | $12 | Prompts ≤200K | source · 2026-10-13 verified | |
| Gemini 3.5 Flash-Lite | $0.3 | $0.03 | $2.5 | source · 2026-10-13 verified | ||
| Grok 4.7 | xAI | $2 | $0.5 | $6 | <200K prompt tokens | source · 2026-10-13 verified |
| Grok Build 0.1 | xAI | $1 | $0.2 | $2 | <200K prompt tokens | source · 2026-10-13 verified |
| Grok 4.3 | xAI | $1.25 | $0.2 | $2.5 | <200K prompt tokens | source · 2026-10-13 verified |
FAQ
How much does prompt caching save?
Cached input is billed at a fraction of the normal input price, often 10x cheaper (e.g. GPT-6.1 Sol $2.00 vs $0.10 per 1M). Savings scale with your cache hit rate.
Which models have cached-input pricing?
Claude Opus 5.5, Claude Opus 5 (legacy), Claude Sonnet 4.6 (legacy), GPT-6 Astra, GPT-6.1 Sol, GPT-6 Luna, GPT-5.6 Sol, GPT-5.3 Codex, Gemini 4 Argon, Gemini 3.8 Flash, Gemini 3.1 Pro Preview, Gemini 3.5 Flash-Lite, Grok 4.7, Grok Build 0.1, Grok 4.3. Models without a listed cache price are excluded.
Does this include cache write costs?
No. Some providers charge extra to write to the cache; this calculator uses list input, cached-input and output prices only.