OpenAI Token Counter (GPT-4, GPT-3.5 & GPT-4o)
Exact token counts for GPT-4, GPT-3.5 and GPT-4o. No upload, no sign-up.
Loading tokenizer…
Paste your text and instantly see how many tokens it uses. Switch between GPT-4 / GPT-3.5 (cl100k_base) and GPT-4o (o200k_base). Your prompt is never uploaded. No sign-up, no credits, no limits.
- Exact API match
- GPT-4 Turbo support
- GPT-4o support
- No upload
- Runs offline
How many tokens is my OpenAI prompt?
Paste your prompt into the box above and pick your model. GPT-4 and GPT-3.5 use the cl100k_base tokenizer. GPT-4o uses o200k_base. Both are counted exactly as the API would count them. Use this to estimate costs or check if your prompt fits inside the context window.
- Price
- Free, no credits, no sign-up
- Tokenizers
- cl100k_base (exact) and o200k_base (exact)
- Models covered
- GPT-4, GPT-4 Turbo, GPT-3.5 Turbo, GPT-4o, GPT-4.1
- Context Windows
- 8k (legacy), 16k (GPT-3.5), 128k (GPT-4/4o)
- Text uploaded to server
- No, runs 100% in your browser
- Input
- Any plain text, code, JSON or chat history
- Extras
- Characters, words, lines, chars-per-token ratio
- Best browsers
- Any modern browser (Chrome, Edge, Firefox, Safari)
- Works offline
- Yes, after the first load
- Not included
- Cost-per-1K calculator, batch file upload, saved history
What is the OpenAI token counter?
This is a free, browser-based tool that counts tokens for OpenAI models. It uses the official `cl100k_base` and `o200k_base` tokenizers, so the counts match the API exactly. Whether you are using older models like GPT-3.5 Turbo or the newer GPT-4o and GPT-4.1, you can check your prompt size with 100% accuracy. It also reports characters, words, lines and the chars-per-token ratio. No file is uploaded, no account is required and the tool works offline after the first load.
It is best for prompt engineering, API cost estimation, system prompt budgeting, and ensuring your long documents fit inside the context window. It is not built for counting image tokens (Vision) or calculating exact dollar amounts, because model prices change often.
How to count OpenAI tokens in 3 steps
- 1
Paste your text
Paste your prompt, system instruction, or conversation into the box. The tool counts as you type.
- 2
Select the model
Choose "GPT-4 / GPT-3.5" for cl100k_base, or "GPT-4o" for o200k_base. The tool uses the exact OpenAI tokenizers for both.
- 3
Read the count
Get the exact token count, plus character and word stats. Verify it fits your context window, then send it to the API.
Features
Exact API Matches
We use the official cl100k_base and o200k_base tokenizers. The numbers you see here match the `usage.prompt_tokens` you get from the OpenAI API exactly. No estimates.
Legacy & Modern Models
Whether you are maintaining an older app on GPT-3.5 Turbo or building a new agent on GPT-4o, this tool covers both tokenizers. GPT-4 Turbo is also fully supported.
Context Window Checks
Check if your prompt fits inside the 8k, 16k, or 128k context windows. Avoid the frustrating "context length exceeded" errors by verifying the count before you send the request.
Instant Character & Word Stats
Get a quick breakdown of characters, words, lines, and the chars-per-token ratio. This is useful for estimating prompt sizes and understanding how efficiently your text is being tokenized.
100% Private
The tokenizer runs locally in your browser using JavaScript. Your proprietary code, confidential data, and prompts are never sent to our servers or stored anywhere.
Zero Friction
No accounts, no email capture, no credit card required. Just open the page, paste your text, and get your exact token count instantly.
What you can use the OpenAI counter for
GPT-3.5 cost optimization
GPT-3.5 uses the cl100k_base tokenizer. If you are running high-volume, low-cost API calls, check the exact token count to ensure your pricing model stays profitable.
GPT-4 Turbo budgeting
GPT-4 Turbo is cheaper than the original GPT-4 but still uses cl100k_base. Count your prompts before running a massive batch job to estimate your total API bill.
Migrating to GPT-4o
Switching from GPT-4 to GPT-4o? Paste your current prompts and switch the dropdown to GPT-4o. The o200k_base tokenizer often uses fewer tokens for the same text, which can save you money.
Checking 16k limits
Many legacy GPT-3.5 models have an 16k token limit. If you are processing long documents, check the count here to avoid API errors.
System prompt optimization
System prompts are sent on every single turn of a conversation. Trimming a 500-token system prompt down to 200 tokens saves 300 tokens on every message.
Code review agents
If you are building an AI coding assistant, code snippets are highly tokenized. Paste your code here to see how much of the context window it actually consumes.
OpenAI Token Counter FAQ
Answers about accuracy, context windows, pricing, and how cl100k_base differs from o200k_base.
Accuracy & Models
Does this cover all OpenAI models?
It covers the most common ones. GPT-3.5 Turbo, GPT-4, and GPT-4 Turbo use cl100k_base. GPT-4o and GPT-4.1 use o200k_base. If a new model uses one of these tokenizers, the count will be exact.
Does this match the OpenAI API exactly?
Yes. We use the official tokenizer libraries. The number you see here will match the `usage.prompt_tokens` returned by the API.
Why do GPT-3.5 and GPT-4 give different counts?
They use different tokenizers. GPT-4 and GPT-3.5 use cl100k_base. GPT-4o uses the newer o200k_base, which is more efficient, especially for non-English text.
Is my prompt saved or logged?
No. The tokenizer runs entirely in your browser. Your text is never sent to our servers, and it disappears when you close or refresh the page.
Context Windows & Cost
What are the context windows for these models?
GPT-3.5 Turbo typically has a 16k window. GPT-4 and GPT-4 Turbo have 128k windows. GPT-4o has a 128k window. The combined input and output must fit inside the window.
How do I calculate my API cost?
Find the current price per 1,000 tokens for your specific model (check OpenAI's pricing page). Multiply your token count by that price. ToolDeck does not display dollar amounts because prices change frequently.
What happens if I exceed the context window?
The API will reject your request. Use this tool to trim your prompt, remove older chat history, or summarize long documents before sending them.
Tokens vs Characters
How many characters is a token?
For plain English, 1 token is roughly 4 characters. Code, math, and non-English languages use more tokens per character. Look at the "Chars / token" stat for the exact ratio.
How many words is 1,000 tokens?
About 750 words of plain English. For code or non-English text, the word count will be lower.
What this tool does well, and where it is limited
This page covers OpenAI models, both legacy and modern. Check this table to see if it fits your use case.
| Use case | Result | Notes |
|---|---|---|
| Counting GPT-4 and GPT-3.5 tokens | Exact | Uses cl100k_base, matching the API exactly. |
| Counting GPT-4o and GPT-4.1 tokens | Exact | Uses o200k_base, matching the API exactly. |
| Counting Claude tokens | Use Claude page | Claude uses a different tokenizer. Use the dedicated Claude Token Counter for approximations. |
| Counting image tokens (Vision) | Not supported | Image tokens are calculated based on resolution and detail level, not plain text. This tool counts text only. |
| Calculating exact dollar cost | Not calculated here | Prices change often and differ between input/output. Multiply the count by the current OpenAI pricing yourself. |
| Counting very large batch files | One at a time | Paste text into the box directly. For large datasets, write a script using the tiktoken library. |
Tips
Switch tokenizers for cost savings
If you are on the fence about which model to use, paste your prompt and switch between GPT-4 and GPT-4o. The o200k_base tokenizer used by GPT-4o is often more efficient, which can lower your per-request cost.
Watch out for whitespace
Newlines and tabs count as tokens. If you are pasting formatted JSON or code, removing unnecessary whitespace can save tokens without losing meaning.
System prompts run every time
A 1,000-token system prompt sent in a 10-turn conversation costs 10,000 tokens. Keep your system prompts as short as possible to optimize costs.
Code is token-heavy
Code uses more tokens than English. If you are building a code analysis tool, always check the exact token count to ensure you do not blow past the context limit.
ToolDeck vs other token counters
Many generic token counters use older tokenizers and give you the wrong number for newer models. ToolDeck uses the exact match for both cl100k_base and o200k_base.
| Feature | ToolDeck | Generic counters |
|---|---|---|
| Uses exact OpenAI tokenizers | Yes, matches API | Often guesses or uses older versions |
| Supports GPT-4 and GPT-3.5 (cl100k) | Yes | Varies by tool |
| Supports GPT-4o (o200k) | Yes | Varies by tool |
| Your text is uploaded to a server | Never | Usually yes |
| Sign-up needed | Not required | Often |
| Free credits or paywalls | None | Often |
| Real-time counting as you type | Yes | Usually yes |
| Character, word and line counts | Yes | Sometimes |
| Works offline after loading | Yes | Usually no |
General comparison. Individual tools differ, and features, limits and pricing change.
Responsible use
The counter is a reference tool. Always confirm the exact count with the provider's API before you depend on it in production. Prices and tokenizers change; this page is updated when they do.
More free AI and developer tools
Model-specific counters and other browser tools. Nothing uploaded, ever.
Token Counter
Count tokens for GPT-4o, GPT-4 and Claude.
GPT-4o Token Counter
Exact o200k_base counting for GPT-4o and GPT-4.1.
OpenAI Token Counter
You are hereGPT-4, GPT-3.5 and GPT-4o tokens.
Claude Token Counter
Estimate Claude 3.5 and Claude 4 tokens.
System Prompt Token Counter
Count system prompt tokens before you send.
Video Editor
Free multi-track editor with AI captions.
Your prompts never leave your device
There is no upload step and no account. The tokenizer runs in your browser, so every count happens on your own device. You can check this yourself: open your browser developer tools, go to the Network tab, then paste text and watch for file uploads. You will not see any. Read how we keep it private.