Hey folks! 👋 I've been knee-deep in ContentBot for a few months now, trying to streamline our content ops, and I think I've finally wrapped my head around their pricing model. It can be a bit confusing at first glance because they use three different units of measurement: **tokens**, **credits**, and **words**. It feels a bit like converting currency while shopping abroad!
After running a bunch of campaigns and analyzing my usage, here’s my practical breakdown of what you're *actually* paying for.
First, let's define the players:
* **Tokens:** This is the fundamental unit AI models (like GPT) actually consume. It's roughly 4 characters for English text. A very rough rule of thumb is that **1000 tokens ≈ 750 words**. ContentBot uses tokens to measure the "cost" of the AI processing your requests.
* **Credits:** This is ContentBot's internal currency. You buy a plan with a monthly credit allowance. Every action you take (generating, rewriting, summarizing) costs a certain number of credits, which is *based* on the token usage.
* **Words:** This is the human-readable output you care about. It's the final article, email, or social post you get. The platform often shows you word counts for clarity.
So, the payment flow works like this: You spend **credits** to cover the **token cost** of generating the **words** you want.
Here’s a concrete example from my workflow:
1. I ask it to "Generate a 500-word blog intro on lead scoring for B2B."
2. The system calculates how many tokens my instruction (the prompt) will use, plus the tokens needed for the 500-word response.
3. That total token count is converted into a credit cost. For instance, a 500-word piece might cost me 12 credits.
4. I get my 500 words, and my monthly credit balance is deducted by 12.
The key takeaways from my experience:
* You're **not directly buying words**. You're buying credits that represent the AI's computational effort (tokens).
* **More complex requests cost more.** A simple "rewrite this paragraph" will chew up far fewer credits than a command to "write a 2000-word comprehensive guide with FAQs and a conclusion."
* **Longer outputs and longer prompts both increase credit cost.** If you paste a huge 1000-word document and ask for a summary, the prompt itself is huge, so it'll be more expensive than a short prompt for a long article.
For someone like me who lives in CRM and forecasting, this model is actually pretty logical once you get it. It's a usage-based cost aligned with the actual AI workload, not just a flat "per word" fee. Just be mindful that asking for multiple variations or doing heavy editing in the same session will add up your token/credit usage faster than just getting a single output.
My advice? Start with a lower plan, run your most common tasks, and check your credit consumption in the dashboard for a week. It’s the best way to forecast what plan you’ll really need.
TIL the hard way that "generate 10 different meta descriptions" can be more credit-intensive than a single, well-crafted blog outline!
Pipeline is king.