Free tools
LLM cost calculator: Claude and OpenAI token prices
Enter the average input and output tokens of a request, your monthly volume and the share of input read from cache, and see per-request and monthly cost for every model in one table. Prices come from the OrqLabs model catalog.
The tool runs entirely in your browser; nothing you enter is sent to a server.
Share of the system prompt and context repeated on every request.
- Per request
- $0.0225
- Per 1,000 requests
- $22.50
- Per month
- $225.00
All models (by monthly cost)
| Model | Input / output $ (1M) | Per request | Per month |
|---|---|---|---|
| Economy | 0.05 / 0.4 | $0.0003 | $3.00 |
| Balanced | 0.25 / 2 | $0.0015 | $15.00 |
| Economy | 0.4 / 1.6 | $0.0016 | $16.00 |
| Economy | 1 / 5 | $0.0045 | $45.00 |
| Flagship | 1.25 / 10 | $0.0075 | $75.00 |
| Balanced | 2 / 8 | $0.008 | $80.00 |
| Balanced | 2 / 10 | $0.009 | $90.00 |
| Balanced | 2.5 / 10 | $0.01 | $100.00 |
| Balanced | 3 / 15 | $0.0135 | $135.00 |
| Flagship · selected | 5 / 25 | $0.0225 | $225.00 |
| Flagship | 5 / 25 | $0.0225 | $225.00 |
| Flagship | 10 / 50 | $0.045 | $450.00 |
Excludes taxes, batch discounts and cache-write cost. Reasoning (thinking) tokens count as output – size the output value accordingly.
How to use it
Pick the model you want to evaluate; the table runs the same assumptions for every other model.
Enter the average input (system prompt + context + user message) and output (the model's answer) tokens of one request.
Enter your monthly request count – for example about 6,000 for 200 chatbot messages a day.
If your system prompt is the same on every request, raise the cache share; input read from cache is billed at a lower rate.
Read the table by monthly cost and choose the cheapest model that still meets your quality bar.
Good to know
Output tokens cost more
Providers price input and output tokens separately, and output is usually several times the input price. For tasks that produce long answers (blog drafts, reports) the output dominates the bill.
Thinking tokens are billed as output
On models with reasoning (thinking) the model's internal reasoning is billed as output tokens too. For simple classification and short replies, economy models or a lower effort setting cut the cost noticeably.
Caching makes repeated context cheap
If the same system prompt and knowledge-base snippets go out with every request, the part read from cache is billed far below the normal input price. On Anthropic models the first cache write costs slightly more than normal input.
How OrqLabs keeps cost in check
OrqLabs computes every run's cost as tokens × catalog price and shows it per model, agent and project on the usage page. Per-workflow and per-agent budget caps and the organization's monthly budget cannot be exceeded. Details: Usage and cost tracking.
Frequently asked questions
Where do the prices come from?
From the OrqLabs model catalog: Anthropic's and OpenAI's list prices in USD per million tokens. Providers can change their prices, so check the provider's pricing page for the exact figure.
How many tokens is one message?
It depends on the language and the model; Turkish text usually takes more tokens than English. Use the provider's token counter, or open the usage details of a test run in OrqLabs.
Are taxes and batch discounts included?
No. The calculation uses list prices in USD before tax; batch discounts and enterprise agreements are not included.
Does OrqLabs add a markup on model usage?
OrqLabs passes model usage through at catalog price and caps it with your monthly budget. On Starter and above you can connect your own Anthropic or OpenAI key, and calls are then billed directly to your provider account.
Other free tools
WhatsApp link generator
Create a wa.me link with a pre-filled message for your website, Instagram bio or email signature.
UTM builder
Build GA4-friendly campaign URLs with UTM parameters, with presets for social media and email.
Social media character counter
Check your copy against the limits of X, LinkedIn, Instagram, Threads, TikTok, YouTube, Facebook and Pinterest.
Get started with OrqLabs
Build your first workflow today
Start on the free plan: connect your accounts, pick a template and stay in control with approval steps. No credit card required.