Track & Report Clawdbot Token Usage Easily

Hey everyone, let's talk about making Clawdbot even better! I've got a feature request that I think will be super helpful for all of us: built-in token usage tracking. Right now, if you're using Clawdbot, you have to jump through a few hoops to figure out how many tokens you're burning through. This new feature aims to solve that problem, giving you a clear view of your API consumption right inside Clawdbot itself. Let's dive into the details, shall we?

The Problem: Token Tracking Headaches

So, what's the deal, guys? Well, currently, there isn't a simple, built-in way to keep tabs on your token usage with Clawdbot. This means you're stuck doing one of two things, neither of which is ideal. You either have to manually check the Anthropic Console (or whatever provider you're using), which can be a bit of a pain, or you have to roll your own custom tracking scripts. That can be a real headache. I mean, let's be real, who wants to spend time writing code just to see how many tokens they've used? Plus, even if you do write a custom script, it might not catch every single interaction, so you might not get a complete picture of your usage.

The Struggles of Manual Tracking

Manual tracking via the Anthropic Console is like, well, it works, but it's not the most convenient thing in the world. You have to switch over to another platform, dig around for the right data, and then try to make sense of it all. It's not exactly a seamless experience. If you're using different channels or models, it becomes even more of a hassle to get a consolidated view of your token usage. And let's not forget the times when the provider dashboards might be down or inaccessible due to some network issues. It just adds another layer of complexity that nobody wants.

Why Custom Scripts Fall Short

Then there's the option of custom tracking scripts. This might seem like a good idea at first, but it can be a real time sink. You have to write the code, debug it, and then make sure it's running correctly. And even then, there's no guarantee that your script will capture every single interaction. Sometimes, interactions can slip through the cracks, leading to inaccurate data. Custom scripts also require maintenance. If the API changes or if you introduce new models, you'll need to update your script. It's just a lot of extra work, especially when you have to balance other priorities. Therefore, a built-in solution would be much more streamlined and efficient, wouldn't you say?

The Solution: A Built-In Token Usage Tracker

Here's what I'm proposing: Let's add native token usage tracking directly into Clawdbot! This would give users a clear and concise way to monitor their API consumption without needing to rely on external dashboards or custom scripts. It's all about making things easier and more efficient, right? Imagine having all your token usage data right at your fingertips, displayed in a neat and easy-to-understand format.

1. Automatic Usage Logging

First things first, we'll need to automatically log token usage after each interaction. This means that every time Clawdbot interacts with an API, we'll keep track of how many tokens are being used. This information would include the number of input tokens (the ones you send to the model) and the number of output tokens (the ones the model generates). We could store this data in a local file, like a JSONL file, or even in a lightweight SQLite database. This would give us a persistent record of all token usage. We'd also include some metadata like the timestamp of the interaction, a session key (so you can track usage per session), the model being used, the channel, and any context tokens, so you get the complete picture of your usage.

2. Built-in Reporting with the /usage Command

Next, we'll add a simple command, /usage, that users can use to view their token usage summary. This would provide an overview of your token consumption over different time periods. Users could use options like /usage 7d to see the last seven days, /usage 30d for the last month, or /usage month for the current month. The report would break down the usage by day and by model, giving users a granular view of their consumption patterns.

3. Optional Cost Estimation

As a bonus, we could even include optional cost estimation. If users configure the model pricing in the config, Clawdbot could calculate and display the estimated costs alongside the token counts. This would give users a more tangible understanding of their spending and help them manage their costs more effectively. This would be a great way to monitor usage and predict the billing, especially for those on metered API plans. This is a very valuable feature.

Example Output

To give you a better idea, here's an example of what the output might look like:

📊 Clawdbot Usage Report (last 7 days)

Total:
  • Interactions: 142
  • Tokens In: 45,230
  • Tokens Out: 128,450
  • Total: 173,680
  • Est. Cost: $2.34

By Model:
  • claude-opus-4-5: 150,000 tokens
  • claude-sonnet-4: 23,680 tokens

By Day:
  • 2026-01-19: 28,400 tokens
  • 2026-01-18: 31,200 tokens
  ...

This would make it super easy to understand your usage at a glance!

Why This Matters

This feature would be particularly useful for several groups of people. First, users on metered API plans, who need to keep a close eye on their costs. Second, those with self-hosted deployments, where provider dashboards might not always be accessible. And third, power users who are always looking for ways to optimize their usage patterns. This feature provides these advantages and more.

Benefits for Metered API Plan Users

For users on metered API plans, the ability to monitor token usage directly within Clawdbot would be a game-changer. It would allow them to track their spending in real-time and make adjustments to their usage patterns to stay within their budget. This feature would give them more control over their expenses.

Benefits for Self-Hosted Deployments

Self-hosted deployments can often lack the same level of monitoring tools available in cloud environments. With a built-in token tracker, self-hosters would have a simple, offline-accessible way to track their usage. This ensures you can monitor and optimize your usage, regardless of your deployment setup.

Benefits for Power Users

Finally, power users who are always striving to optimize their workflows would love this feature. By analyzing the token usage data, they can identify areas where they can improve their prompts, adjust their models, or optimize their overall interaction patterns, leading to more efficient and cost-effective usage.

Alternatives Considered

We touched on the alternatives earlier. Manual tracking via custom scripts, while functional, can miss interactions. Relying solely on provider dashboards lacks local/offline access and per-channel breakdowns. The proposed solution offers a comprehensive and integrated approach.

Conclusion

Adding built-in token usage tracking to Clawdbot would be a win-win for everyone. It simplifies usage monitoring, provides valuable insights, and ultimately helps users manage their costs and optimize their usage patterns. What do you all think? Let me know your thoughts in the comments! I think this will be an extremely positive addition to the bot. I cannot wait to see what you guys think, and I hope you like the idea.