API cost calculator · Prices retrieved 2026-11-11
Estimate monthly Gemini 4 Carbon API costs at the introductory price — through November 30, 2026 — and see what the same usage costs at the regular price starting December 1. Then compare Gemini 4 Carbon vs Claude Fable 5.2, including Fast mode and cache pricing. No login, no server requests: the math runs in your browser.
Carbon · intro
through 2026-11-30
$2.00/$10.00
input / output per 1M tokens
Carbon · regular
from 2026-12-01
$4.00/$20.00
input / output per 1M tokens
Claude Fable 5.2
standard
$8.00/$40.00
input / output per 1M tokens
Instant estimate · Rates retrieved 2026-11-11
Enter your expected monthly token volume. Totals update as you type or drag. Rates are in USD per 1 million tokens. Cached tokens are counted separately from uncached input so they are not billed twice.
Gemini 4 Carbon · Introductory · Through November 30, 2026
Estimated monthly cost
Carbon intro $91.00 per month
USD per month at the introductory price
From 2026-12-01, the same Gemini 4 Carbon usage costs 2.00× the introductory price — $91.00 more per month. Introductory pricing ends 2026-11-30.
Gemini 4 Carbon at the introductory price is $271.00 (75% less) lower than Claude Fable 5.2 standard for this mix. This is a price comparison only, not a quality or capability claim.
Each cell shows the monthly USD total and the per 1M token rate used. Fast mode cache rates are derived from official Fast mode input pricing multiplied by Anthropic's published cache multipliers.
| Line item | Monthly tokens | Carbon intro | Carbon regular | Fable 5.2 | Fable 5.2 Fast |
|---|---|---|---|---|---|
| Uncached input | 20M | $40.00 $2.00 / 1M tokens | $80.00 $4.00 / 1M tokens | $160.00 $8.00 / 1M tokens | $320.00 $16.00 / 1M tokens |
| Cached input (implicit cache hits) | 10M | $1.00 $0.10 / 1M tokens | $2.00 $0.20 / 1M tokens | $2.00 $0.20 / 1M tokens | $4.00 $0.40 / 1M tokens (derived) |
| Cache writes | 0 | Not billed. Gemini 4 Carbon supports implicit caching only; explicit cache writes are not billed because they are not supported. | Not billed. Gemini 4 Carbon supports implicit caching only; explicit cache writes are not billed because they are not supported. | $0.00 $10.00 / 1M tokens | $0.00 $20.00 / 1M tokens (derived) |
| Output (includes thinking tokens) | 5M | $50.00 $10.00 / 1M tokens | $100.00 $20.00 / 1M tokens | $200.00 $40.00 / 1M tokens | $400.00 $80.00 / 1M tokens |
| Monthly total | — | $91.00 | $182.00 | $362.00 | $724.00 |
Official list rates
These are the paid-tier list rates this calculator uses, copied from public vendor pages on 2026-11-11. Units are USD per 1 million tokens.
| Gemini 4 Carbon | Intro through 2026-11-30 | Regular from 2026-12-01 |
|---|---|---|
| Uncached input | $2.00 | $4.00 |
| Output, including thinking | $10.00 | $20.00 |
| Cached input (implicit) | $0.10 | $0.20 |
| Explicit cache writes | Not supported | |
| Claude Fable 5.2 | Standard | Fast mode |
|---|---|---|
| Uncached input | $8.00 | $16.00 |
| Output | $40.00 | $80.00 |
| Cache reads | $0.20 | $0.40 (derived) |
| 5-minute cache writes | $10.00 | $20.00 (derived) |
| 1-hour cache writes | $16.00 | $32.00 (derived) |
Sources: aistudio.google.com/docs/pricing, Gemini 4 Carbon announcement, platform.claude.com pricing, and Fable 5.2 what's new. Fast mode cache rates are derived from official Fast mode input prices times official cache multipliers (1.25×, 2×, and 0.025×).
FAQ
Google's public pricing lists a paid-tier introductory window through 2026-11-30: $2.00 per 1 million input tokens, $10.00 per 1 million output tokens (including thinking tokens), and $0.10 per 1 million implicit cache-hit tokens. Starting 2026-12-01, those rates become $4.00, $20.00, and $0.20. The same introductory rates were also announced in the Gemini 4 Carbon launch post. This calculator shows both monthly totals for the usage you enter.
Enter monthly uncached input tokens, output tokens, and cached input tokens as separate fields so cached tokens are not billed twice. For Gemini 4 Carbon, put thinking tokens in the output field — official pricing says output includes thinking tokens. Cached input should be implicit cache hits for Gemini, or cache reads (hits and refreshes) for Claude. Cache-write tokens are optional and apply only to Claude Fable 5.2 prompt-cache writes. Gemini 4 Carbon supports implicit caching only; explicit cache writes are not supported, so that field does not change the Carbon totals.
This page compares list prices, not model quality. Claude Fable 5.2 standard rates are $8 per 1 million uncached input tokens and $40 per 1 million output tokens, with cache reads at $0.20 and cache writes at $10 (5-minute) or $16 (1-hour). Fast mode is $16 input and $80 output per 1 million tokens; Fast mode cache rates used here are derived by applying Anthropic's published cache multipliers to that Fast mode input price. For many input/output mixes, Carbon's introductory and even regular list prices are lower than Fable 5.2 standard. Actual bills also depend on thinking volume, cache hit rate, tools, and whether you use Fast mode.
No. This is an unofficial, client-side estimate. Prices were copied from vendor public pricing pages on 2026-11-11 and are not an official quote from Google or Anthropic. Vendors can change rates, and your invoice may include taxes, committed-use discounts, grounding fees, batch discounts, or regional premiums that this page does not model. Always check the official Gemini API pricing page and Anthropic pricing page before you spend.
Official references: Gemini API pricing, Carbon launch post, and Claude pricing.