Agent reports and credit usage

Overview

Once an agent is running, you will want to know two things: what did it do, and what did it cost. Every agent in WorkLLM has its own Results page, and the My Executions tab shows all your agents in one place. Credits work the same way as in the rest of WorkLLM.

This article shows where to find your results and logs, and how credits are counted for agents, including what does not use credits.

Seeing results and reports

Every agent has its own Results page. Here you can see what the agent has done and what it produced.

The Results page for one agent

On the Results page, you can see:

• The result of each run.
• Who ran the agent.
• How many times the agent has run.
• How many credits each run used.
• The logs for each run.

Open the logs of a run to see the full details of what happened during that execution.

My Executions: all your agents in one place

On the left side of the screen, you will find the My Executions tab. It shows your executions for all agents in a single table. You can see which agent ran, and when it ran.

This is especially helpful when you have agents that start on their own, for example on a schedule or when something happens in an app. With My Executions, you can quickly check:

• How many agents ran today.
• How many credits they used.
• What happened in each run, by opening its full execution log.

Tip: If an agent that starts on its own behaves in an unexpected way, open its execution log from My Executions. The log shows the details of the run, which makes it easier to find the cause.

Managing credits for agents

Agents use the same credits as the rest of WorkLLM. Nothing separate to learn.

What	Credits
Connecting an app	10 credits, one time
Creating and testing an agent	Credits for the AI work involved
Running an agent	Credits for the model work in each step, which depends on the model and how much text it reads and writes
Web search	1 to 3 credits per search, depending on the search
Deep research and images	An extra charge, on top of the step
Saving a file to memory	Credits once, to process it

What does not use credits

A few things do not use WorkLLM credits at all:

• Your own API key (BYOK). When you use a model on your own API key, WorkLLM does not charge credits for the answer itself. Some other parts of WorkLLM still use credits, because they do not run on your key.
• Searching your saved memory. Looking up a file you saved to memory has no separate charge. You pay for the answer only.
• Failed requests. A request that fails should not use credits. If you were charged for one, email [e-mail] and we will refund the credits.

What still uses credits when you use your own key

If you use your own API key for a model, the model itself does not use WorkLLM credits. Memory, web search, and connecting an app still do. For more understanding please read How WorkLLM credits work. 

Frequently asked questions

• Where can I see what my agent produced?
  Open the agent and go to its Results page. It shows the result of each run, who ran it, and the logs.
• How can I see how many credits a run used?
  The Results page shows the credits used by each run. My Executions shows the credits used by your agents in one table.
• What is My Executions?
  It is a tab on the left side of the screen. It shows your executions for all agents in a single table, so you can see which agent ran and when, and open the full execution log of any run.
• What happens if a step fails?
  You can see the failure in the agent’s execution log. Open the log, check the details, and then edit the agent to fix the cause. For how to edit, see Editing and deleting agents.
• Do agents use credits if I bring my own API key?
  The model itself does not. When a model runs on your own key, your provider bills you directly and WorkLLM does not charge credits for those answers. A few things still use credits, because they do not run on your key: memory, web search, document processing (embeddings), and connecting an app. If you use deep research or image generation because your own model cannot do them, those use credits too. WorkLLM never switches between your key and your credits on its own. The model you pick decides which one pays.
• My credits are dropping faster than expected. What should I check?
  Open the Credit Usage report and look at usage by agent and model. The usual causes are steps on higher-end models, web search, deep research or image generation, and large files attached each run. Switch simple steps to lighter models, save repeated files to memory, and run the agent less often if you can.
• Does a failed request use credits?
  A request that fails should not use credits. If you were charged for one, email [e-mail] and we will refund the credits.