Bill Projector

What will this cost me to run?

Enter your expected traffic and see the monthly bill across every model — before a single user signs up.

Start from

BudgetQwen

Qwen3.7 Flash

$0.03 in · $0.13 out / 1M

$119

per month · $0.02 / user

StandardxAI

Grok Build 0.1

$1 in · $2 out / 1M

$2,568

per month · $0.51 / user

PremiumAnthropic

Claude Fable 5 (batch)

$5 in · $25 out / 1M

$21,570

per month · $4.31 / user

#ModelIn / OutRelativeMonthly

Qwen3.7 FlashCHEAPEST

$0.03/$0.13

$119

Qwen3.5-9B

$0.1/$0.15

$221

Qwen3.5-Flash

$0.065/$0.26

$242

Gemma 4 26B A4B

$0.07/$0.34

$296

GLM 4.7 Flash

$0.06/$0.4

$321

DeepSeek V4 Flash

$0.14/$0.28

$360

Gemma 4 31B

$0.14/$0.4

$424

GPT-5.4 Nano (batch)

$0.1/$0.625

$506

Mistral Small 4

$0.15/$0.6

$557

Gemini 3.1 Flash Lite (batch)

$0.125/$0.75

$614

Qwen3.5-35B-A3B

$0.14/$1

$784

Qwen3.6 35B A3B

$0.14/$1

$784

Qwen3 Coder Next

$0.18/$0.9

$786

Qwen3.6 Flash

$0.188/$1.125

$922

Gemini 3.5 Flash Lite (batch)

$0.15/$1.25

$947

GPT-5.4 Nano

$0.2/$1.25

$1,013

DeepSeek V4 Pro

$0.435/$0.87

$1,072

Qwen3.5-27B

$0.195/$1.56

$1,193

Qwen3.7 Plus

$0.32/$1.28

$1,206

Gemini 3.1 Flash Lite

$0.25/$1.5

$1,229

Gemini 3.1 Flash Lite Preview

$0.25/$1.5

$1,229

Qwen3.5 Plus 2026-02-15

$0.26/$1.56

$1,278

Qwen3.5 Plus 2026-04-20

$0.3/$1.8

$1,474

Qwen3.5-122B-A10B

$0.26/$2.08

$1,590

Qwen3.6 Plus

$0.325/$1.95

$1,597

Qwen3.6 27B

$0.3/$2

$1,659

GPT-5.4 Mini (batch)

$0.375/$2.25

$1,842

Gemini 3.5 Flash Lite

$0.3/$2.5

$1,894

Qwen3.5 397B A17B

$0.39/$2.34

$1,916

GLM 5.2

$0.749/$2.354

$2,431

GPT-5.6 Luna

$0.5/$3

$2,457

GPT-5.6 Luna Pro

$0.5/$3

$2,457

Kimi K2.5

$0.57/$2.85

$2,480

Kimi K2.6

$0.646/$2.72

$2,505

Grok Build 0.1

$1/$2

$2,568

GLM 5

$0.95/$2.55

$2,835

Kimi K2.7 Code

$0.73/$3.5

$3,101

GLM 5.1

$0.966/$3.036

$3,135

Grok 4.20

$1.25/$2.5

$3,183

Grok 4.20 Multi-Agent

$1.25/$2.5

$3,183

Grok 4.3

$1.25/$2.5

$3,183

Gemini 3.6 Flash (batch)

$0.75/$3.75

$3,236

Qwen3 Max Thinking

$0.78/$3.9

$3,365

Gemini 3.5 Flash (batch)

$0.75/$4.5

$3,686

GPT-5.4 Mini

$0.75/$4.5

$3,686

GLM 5 Turbo

$1.2/$4

$4,042

GLM 5V Turbo

$1.2/$4

$4,042

Claude Sonnet 5 (batch)

$1/$5

$4,314

Qwen3.7 Max

$1.475/$4.425

$4,673

Gemini 3.1 Pro Preview (batch)

$1/$6

$4,914

Qwen3.6 Max Preview

$1.027/$6.162

$5,047

GPT-5.4 (batch)

$1.25/$7.5

$6,143

GPT-5.6 Terra

$1.25/$7.5

$6,143

GPT-5.6 Terra Pro

$1.25/$7.5

$6,143

Grok 4.5

$2/$6

$6,282

Gemini 3.6 Flash

$1.5/$7.5

$6,471

Mistral Medium 3.5

$1.5/$7.5

$6,471

Gemini 3.5 Flash

$1.5/$9

$7,371

Claude Sonnet 5

$2/$10

$8,628

Gemini 3.1 Pro Preview

$2/$12

$9,828

Gemini 3.1 Pro Preview Custom Tools

$2/$12

$9,828

GPT-5.2-Codex

$1.75/$14

$10,700

GPT-5.3 Chat

$1.75/$14

$10,700

GPT-5.3-Codex

$1.75/$14

$10,700

Claude Opus 4.6 (batch)

$2.5/$12.5

$10,785

Claude Opus 4.7 (batch)

$2.5/$12.5

$10,785

Claude Opus 4.8 (batch)

$2.5/$12.5

$10,785

GPT-5.4

$2.5/$15

$12,285

GPT-5.5 (batch)

$2.5/$15

$12,285

Claude Sonnet 4.6

$3/$15

$12,942

Kimi K3

$3/$15

$12,942

Claude Fable 5 (batch)

$5/$25

$21,570

Claude Opus 4.6

$5/$25

$21,570

Claude Opus 4.7

$5/$25

$21,570

Claude Opus 4.8

$5/$25

$21,570

Claude Opus 5

$5/$25

$21,570

GPT Chat Latest

$5/$30

$24,570

GPT-5.5

$5/$30

$24,570

GPT-5.6 Sol

$5/$30

$24,570

GPT-5.6 Sol Pro

$5/$30

$24,570

Claude Fable 5

$10/$50

$43,140

Claude Opus 4.8 (Fast)

$10/$50

$43,140

Claude Opus 5 (Fast)

$10/$50

$43,140

Claude Opus 4.7 (Fast)

$30/$150

$129,420

GPT-5.4 Pro

$30/$180

$147,420

GPT-5.5 Pro

$30/$180

$147,420

Assumes 30 days/month. Cached input billed at each model's own cached rate. See full rates on the pricing page.

How the calculator works

The projection is a straightforward token-volume model. From your inputs we derive monthly request volume and token counts, then price them against every model’s live rates:

Prompts per month = monthly active users × prompts per user per day × 30 days.
Input tokens = prompts per month × average input tokens per prompt.
Output tokens = prompts per month × average output tokens per prompt.
Monthly cost = (input tokens ÷ 1M) × blended input rate + (output tokens ÷ 1M) × output rate, where the blended input rate applies your prompt-cache hit rate: cached tokens are billed at each model’s own cached-input rate and the rest at the standard input rate.

In other words, a 30% cache hit rate means 30% of your input tokens are billed at the model’s cached rate and 70% at the full input rate. Output tokens are never cached, so they’re always billed at the standard output rate.

A worked example

Say you have 5,000 monthly active users, each sending 8 prompts a day, with 1,500 input tokens and 500 output tokens per prompt, at a 30% cache hit rate. That’s 1.2M prompts/month → 1.8B input tokens and 600M output tokens. For a model priced at $2 input / $10 output / $0.20 cached per 1M, the blended input rate is ($2 × 0.7) + ($0.20 × 0.3) = $1.46/1M, so input ≈ 1,800 × $1.46 = $2,628 and output ≈ 600 × $10 = $6,000 — about $8,628/month. Change any slider and the whole table re-ranks instantly.

If this estimate looks higher than you'd like, four concrete levers move it the most — caching, batching, model routing, and context trimming — each with worked math in How to Cut Your LLM API Bill by 80%.

The estimate assumes a 30-day month and steady traffic; real bills vary with burst patterns, retries, and tool-use overhead. Treat it as a planning figure and confirm live rates on the pricing page. See methodology for data sources.

Frequently asked questions

Is the calculator accurate for my real usage?

It is a planning estimate. It models steady monthly traffic at a 30-day month using live per-token rates. Real bills vary with traffic bursts, retries, tool-use overhead, and streaming, so treat the figure as a well-grounded projection rather than an invoice.

How is the prompt-cache hit rate applied?

Only to input tokens. At a 30% hit rate, 30% of input tokens are billed at the model’s cached-input rate and 70% at the standard input rate. Output tokens are never cached and are always billed at the full output rate.

Are the values I enter sent anywhere?

No. All calculation happens in your browser. The traffic and token numbers you type are never transmitted to or stored on our servers.

Why does the cheapest model change when I adjust output length?

Because models have very different input-to-output price ratios. A model with cheap input but expensive output can win for short answers and lose for long ones, so the ranking re-sorts as you change the input/output mix.

What counts as a 'prompt'?

One request/response round trip to the model. Multi-turn conversations send the prior history as input each turn, so long chats increase input tokens — account for that in your average input-tokens figure.