# Capacity you offer on Golem is not a paid task

Cores, memory, and disk are capacity you are willing to share. Task counts in status and GLM on the monitor are separate. Capacity or an online flag does not mean a task was paid.

- Canonical: https://iotahome.site/en/learn/golem-offers-tasks-availability
- Publisher: IOTA Watch
- Topics: Golem
- Published: 2026-10-24
- Updated: 2026-10-24

## At a glance

Read the resources you offer, then task counts in status, and only then whether the GLM record has a number.

## Capacity, tasks, and GLM

| What you see | Question it answers | Question it does not answer |
| --- | --- | --- |
| Cores / memory / disk | How many threads, how much RAM, and how much disk you offer | Whether a requestor has paid for that capacity |
| Last hour in progress / processed | Whether tasks are running and how many finished in that hour | The GLM price of each task or the amount credited |
| Total processed, including failures | How much work was attempted, failures included in that wording | How many attempts were paid |
| Online flag, or providers computing network-wide | Whether your record is marked online, or how many providers are computing | GLM for your machine |

Publishing CPU threads, memory, and disk on Golem only says you are willing to share those resources. A requestor actually sending work, and that work turning into GLM, are the next two steps. Capacity, task counts, and payment do not stand in for one another.

## What each item means

In the install guide, YA_RT_CORES, YA_RT_MEM, and YA_RT_STORAGE are the threads, RAM in GiB, and disk in GiB you share. The image ships with small defaults. The guide tells you to set what you really want to share and to leave headroom for the host. The initial default price is stated as 0.1 GLM per hour per thread, and the same page says that can be high for the current market, so you compare the provider list on Stats and set your own price. The price lives in provider data. It is not income on the monitor.

The Tasks block in golemsp status shows processed and in progress for the last hour, plus total processed, and it says failures are included. Those are work counts, not GLM. The Golem Stats front page also counts providers computing for requestors across the network. That is a network scale, not your machine’s bill.

The monitor can show node name, CPU model, core count, the online flag, and rolling 24-hour plus lifetime GLM. It does not list each task.

## What to inspect

- Read the cores, memory, and disk you actually offer in status or in the install file. Do not substitute the whole machine.

- See whether last-hour in progress and processed are zero. Zero means that status sample had no matching tasks. It does not mean capacity equals income.

- Total processed includes failures. A failure count is not a count of payments received.

- Back on the monitor, a GLM column with no number stays unknown. Do not invent income because the core count is above zero.

## What to do next

- Capacity is large but task counts stay at zero: start with [offline and reachability checks](https://iotahome.site/en/learn/golem-provider-offline-checks), including the virtualization device and the network. A high price can also make requestors pass you by. The monitor cannot prove that.

- Task counts rise while GLM is still empty: payment and balances are in [GLM payments and capacity](https://iotahome.site/en/learn/golem-glm-payments-vs-capacity). Do not multiply task counts by the default 0.1 to estimate a payout.

- GLM is present but looks unrelated to the cores you offer: cores are not hours worked. Read the two GLM figures in [rolling 24 hours and lifetime](https://iotahome.site/en/learn/golem-monitor-guide).

## Example

Illustrative, not a measurement: the file offers 4 threads, 8 GiB of RAM, and 40 GiB of disk, and status shows 0 processed and 0 in progress in the last hour, with total processed 0. That sample had no task work. It does not set rolling 24-hour GLM to 0. If that column is missing, it stays unknown. The network-wide provider count on the chart is not your income either.

## Related tool

On [the Golem project page](https://iotahome.site/en/projects/golem), save only the node you want to watch. Free is five saved devices per project, Pro 50. Extra public fields on the same machine do not take another slot.

## Common questions

### Is the default 0.1 GLM a rate I have already earned?

No. It is the initial per-thread hourly price in the install guide, with a warning that it may be high. It is not income on the monitor.

### Can the network-wide computing count stand in for my tasks?

No. That is the network scale on Stats. Your own counts are the Tasks block on the machine.

### If the node is online and cores are above zero, is income unknown or zero?

When the GLM column has no number, it is unknown. It is zero only when the source explicitly shows zero. Online and core count do not fill that blank.

## Sources and implementation

- [Golem Network](https://golem.network/)
- [Golem provider installation](https://docs.golem.network/docs/providers/provider-installation)
- [Golem Stats](https://stats.golem.network/)