> ## Documentation Index
> Fetch the complete documentation index at: https://docs.median.sh/llms.txt
> Use this file to discover all available pages before exploring further.

# Pricing

> Plans, credits, and how usage is billed.

Median runs on a **credit-based model**. Every dollar you pay becomes a
dollar of credits — and you spend them as you use the product. AI usage is
charged at cost, and integration events draw a flat per-event price.

## Plans

| Plan    | Price | Credits / mo | Trial  | Extras                               |
| ------- | ----- | ------------ | ------ | ------------------------------------ |
| Free    | \$0   | \$0.50       | —      | Standard AI model only · hard-capped |
| Starter | \$5   | \$5          | 7 days | Advanced AI model                    |
| Plus    | \$10  | \$10         | —      | Advanced AI model                    |
| Scale   | \$20  | \$20         | —      | Advanced AI model · Priority support |

Paid plans unlock every feature, including the advanced AI model. The only
difference between paid plans is the amount of credits granted each cycle.

### Free tier caveats

The Free plan is meant for evaluating Median end-to-end without a card. It
has a few hard limits that paid plans don't:

* **Standard AI model only** — feedback extraction and task generation run
  on the lower-cost model. The advanced model is reserved for paid plans.
* **Hard-capped at \$0.50 of credits per month** — usage stops cleanly when
  credits run out; nothing is billed beyond that.
* **No paid overages** — the "Hard cap usage at plan credits" toggle is
  permanently on. To exceed \$0.50 in a month, upgrade to a paid plan.

## How credits are spent

Credits are a single shared pool. Both AI and integration events draw from
it.

| Activity                                                       | Cost          |
| -------------------------------------------------------------- | ------------- |
| Integration event ingested (Discord, Slack, GitHub, Linear, X) | \$0.007       |
| AI task generation, summaries, replies                         | At model cost |

### AI pricing

AI is billed at the underlying model rate — no markup. Median routes most
work to a **fast, low-cost model** and reserves a **larger, more capable
model** for heavier reasoning tasks. Most everyday actions cost a fraction
of a cent.

## Overages

On paid plans, usage beyond your monthly credits is **auto-charged at the
end of the billing cycle** at the same per-event and per-token rates. There
is no overage premium — \$1 of usage costs \$1 whether it is inside or
outside your included credits.

The Free plan does not support paid overages. Once your \$0.50 of credits
is spent, ingest and AI generation pause until the next cycle or an
upgrade.

### Hard-cap mode

You can hard-cap usage at your plan's credits. With this enabled:

* AI generation pauses once credits are exhausted
* Integrations stop ingesting new events
* The task board stays available — Linear, GitHub, and Discord syncs
  resume on the next cycle or after an upgrade

Toggle this from **Billing → "Hard cap usage at plan credits"** in the
dashboard.

## Worked examples

<AccordionGroup>
  <Accordion title="Trying Median, free plan">
    A developer signs up to evaluate Median, ingests \~30 Discord messages
    and runs a handful of AI task generations.

    * Events: 30 × \$0.007 = **\$0.21**
    * AI: \~\$0.10 (standard model only)
    * **Total used: \~\$0.31** of \$0.50 — well within the free cap.
  </Accordion>

  <Accordion title="Solo developer, $5 starter plan">
    A solo developer ingests \~300 GitHub events and runs \~50 AI task
    generations in a month.

    * Events: 300 × \$0.007 = **\$2.10**
    * AI: \~\$1.20
    * **Total used: \~\$3.30** of \$5 in credits — no overage.
  </Accordion>

  <Accordion title="Small team, $10 plus plan">
    A team ingests \~1,000 events across Discord and Linear, with \~150 AI
    runs.

    * Events: 1,000 × \$0.007 = **\$7.00**
    * AI: \~\$3.50
    * **Total used: \~\$10.50** — \$0.50 overage, billed at cycle end.
  </Accordion>

  <Accordion title="Power user, $20 scale plan">
    Heavy GitHub + Linear ingest (\~2,500 events) with \~400 AI runs.

    * Events: 2,500 × \$0.007 = **\$17.50**
    * AI: \~\$2.40
    * **Total used: \~\$19.90** of \$20 — no overage. Priority support
      included.
  </Accordion>
</AccordionGroup>

## FAQ

<AccordionGroup>
  <Accordion title="Do unused credits roll over?">
    No. Credits reset at the start of each billing cycle.
  </Accordion>

  <Accordion title="Can I change plans mid-cycle?">
    Yes. Upgrades take effect immediately and are pro-rated. Downgrades
    schedule for the next cycle.
  </Accordion>

  <Accordion title="Can I stay on the free plan forever?">
    Yes. The free plan has no time limit — it just hard-caps at \$0.50 in
    credits per month and routes AI work to the standard model.
  </Accordion>

  <Accordion title="What counts as an integration event?">
    Anything ingested from a connected source — a new Discord message in a
    feedback channel, a GitHub issue, a Linear ticket, a Slack message in a
    feedback channel, or a tweet.
  </Accordion>

  <Accordion title="Where do I see my usage?">
    The [Billing page](https://median.sh/app/billing) shows current
    credits, daily consumption, AI vs event breakdown, and any overages.
  </Accordion>
</AccordionGroup>