Which AI models are included?

We include GPT-4o, GPT-4o-mini, GPT-4, o1, o3-mini, Claude 3.5 Sonnet, Claude 3.5 Haiku, Gemini 2.0 Flash, Gemini 2.0 Pro, Gemini 1.5 Pro, DeepSeek-V2, DeepSeek-R1, Mistral Large, Mistral Small, Llama 3.1 405B, Command R+, and Grok-2.

What does output ratio mean?

The output ratio is how many output tokens you expect relative to input. For example, a ratio of 0.5 means you expect half as many output tokens as input tokens. Typical use cases: summarization (0.2-0.3), Q&A (0.5-1.0), code generation (1.0-2.0).

✦ Free AI Tool

AI Token Cost Calculator

Paste any text to estimate token usage and compare input/output costs across all major AI models. See exactly what each provider charges.

📝 Paste Your Text (or type below)

Words: 0 Characters: 0 Estimated Tokens: 0

🔧 Output Settings

Output / Input Ratio (how many output tokens per input token)

0.5

Summarization (0.2) Q&A (0.5) Code Gen (1.5+)

📊 Output Tokens (auto-calculated)

Output Tokens

Total Tokens

💰 Cost Comparison — All Major AI Models

Model	Provider	Input / 1M tok	Output / 1M tok	Your Cost

Prices shown in USD. Updated May 2026. Always verify with provider.

🧠 How Token Estimation Works

Tokens are chunks of text that AI models process. A rough rule of thumb:

📝

1 Token ≈ 4 Characters

In English, one token is roughly 4 characters including spaces. "Hello, world!" = about 3 tokens.

"The quick brown fox" → 4 tokens

📊

1 Token ≈ 0.75 Words

Most English text averages 0.7–0.8 words per token. This tool uses the character-based estimate.

100 words ≈ 133 tokens

⚡

Pricing Is Per Token

Providers charge per 1M tokens. A 10¢/1M rate means 10,000 tokens costs just 1 cent.

DeepSeek: $0.14/1M input tokens

Actual token counts vary by provider (OpenAI uses a different tokenizer than Anthropic or Google). This tool gives a close estimate for cost comparison purposes.

📋 Typical Use Case Costs

See how much common AI tasks cost across different providers.

📄

Summarize a Page

500 word article → ~670 input tokens, ~200 output tokens (ratio 0.3)

GPT-4o: ~$0.004 | DeepSeek: ~$0.0001

💬

Chat Conversation

1,000 word chat history + query → ~1,330 input tokens, ~500 output (ratio 0.38)

Claude Sonnet: ~$0.011 | Gemini Flash: ~$0.0003

📝

Write a Blog Post

200 word prompt → ~270 input tokens, ~1,000 output words (ratio ~5)

GPT-4o: ~$0.011 | DeepSeek: ~$0.0003

❓ Frequently Asked Questions

How do you estimate tokens from text? ▾

We use a rough estimate: 1 token ≈ 4 characters in English text, or about 0.75 words per token. Actual token counts vary by model tokenizer, but this gives a close approximation for estimating costs. Different providers use different tokenization algorithms — OpenAI uses tiktoken, Anthropic uses their own, etc.

What's the cheapest AI model? ▾

Gemini 2.0 Flash ($0.10/1M input) and DeepSeek-V2 ($0.14/1M input) are the cheapest among major providers. GPT-4o-mini ($0.15/1M input) is OpenAI's budget option. For output-heavy tasks, DeepSeek-V2 ($0.28/1M output) and Gemini Flash ($0.40/1M output) lead on price.

Why does output cost more than input? ▾

Generating tokens (output) requires more compute than processing existing tokens (input). The model must predict each token sequentially, which is compute-intensive. Most providers charge 2–5x more for output tokens than input tokens.

Is my text sent to any server? ▾

No. All processing happens in your browser using JavaScript. Your text never leaves your device. The token estimation and cost calculations are 100% client-side.

What is DeepSeek's 10M token rate? ▾

DeepSeek-V2 charges $0.14 per 1M input tokens, which means 10 million input tokens cost just $1.40. For output tokens, it's $0.28 per 1M, so 10M output tokens cost $2.80. This is among the cheapest rates in the industry, making DeepSeek extremely popular for high-volume applications.