> For the complete documentation index, see [llms.txt](https://docs.crew.you/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.crew.you/billing-and-plans/credits-explained.md).

# Credits Explained

Credits are Crew's usage currency. Every AI interaction consumes credits based on the complexity of the conversation and the AI model used.

## How Credits Are Consumed

When you send a message, Crew reads your input, thinks about it, potentially uses tools, and generates a response. The credit cost depends on:

* **Which AI model** handles your request
* **Input tokens**: Your message + conversation history + tool results
* **Output tokens**: Crew's response text
* **Number of tool calls**: Each tool use adds tokens

### Credit Cost by Model

| Model                | Credits per Turn (typical) | Best For                                   |
| -------------------- | -------------------------- | ------------------------------------------ |
| **Lightning**        | \~4                        | Quick answers, greetings, simple questions |
| **Core**             | \~6                        | Everyday tasks, email, calendar            |
| **Insight**          | \~32                       | Analysis, coding, document creation        |
| **Prime** (Pro only) | \~131                      | Maximum capability tasks                   |

### Typical Task Costs

| Action                              | Approximate Credits |
| ----------------------------------- | ------------------- |
| Simple Q\&A (Lightning)             | 3–6                 |
| Email read + summary (Core)         | 10–20               |
| Web search + summary (Core/Insight) | 15–30               |
| Document creation (Insight)         | 30–50               |
| Multi-tool complex task (Insight)   | 50–100              |
| Code execution + analysis (Insight) | 40–80               |
| Driving Mode TTS (per response)     | 1–120               |

### Per-Token Rates

| Model     | Input (per 1K tokens) | Output (per 1K tokens) |
| --------- | --------------------- | ---------------------- |
| Lightning | 0.75 credits          | 4.5 credits            |
| Core      | 0.8 credits           | 4 credits              |
| Insight   | 4.5 credits           | 22.5 credits           |
| Prime     | 18.75 credits         | 93.75 credits          |

## Monthly Allowance

Credits reset at the start of each billing cycle:

| Plan  | Monthly Credits |
| ----- | --------------- |
| Free  | 4,000           |
| Basic | 8,000           |
| Plus  | 25,000          |
| Pro   | 60,000          |

## When Credits Run Out

When your monthly credits are depleted, you can still **keep chatting for free** — Crew automatically switches to the **Core** model. Your conversations continue without interruption, just with a simpler model.

To resume using more powerful models (Insight, Prime):

* Wait for your monthly reset
* Purchase a credit pack (paid plans only)
* Upgrade your plan

## Credit Packs

Buy additional credits anytime (paid plans only):

| Pack     | Credits | Price |
| -------- | ------- | ----- |
| Starter  | 2,000   | $5    |
| Standard | 5,000   | $10   |
| Premium  | 15,000  | $25   |

Purchased credits are added as **bonus credits** that carry over month-to-month — they never expire with your billing cycle.

## Auto-Recharge

Set up automatic credit purchases so you never run out:

1. Go to **Settings → Billing**
2. Enable **Auto-Recharge**
3. Set a **threshold** (100, 200, or 500 credits)
4. Choose a **credit pack**

When your balance drops below the threshold, the selected pack is purchased automatically.

## Tracking Usage

### Credit Bar

The sidebar shows a visual credit bar:

* **Green**: Plenty remaining
* **Orange**: Getting low (< 30%)
* **Red**: Almost depleted (< 10%)

### Transaction History

Go to **Settings → Billing → Transactions** for detailed usage:

* Every conversation turn with credit cost
* Model used and token breakdown (input, output, cache)
* Credit purchases and plan changes
* Referral bonuses
* Mission credit usage
* Filter by type: Usage, Purchases, Plan Changes, Resets

Transactions are grouped by session — multiple turns in the same conversation within 2 minutes are bundled for easy reading.

## Tips to Optimize Credit Usage

1. **Be concise**: Shorter messages use fewer input tokens
2. **Be specific**: "Find emails from John about budget" is cheaper than "check all my emails"
3. **Start new conversations**: For new topics, start fresh instead of carrying long history
4. **Use scheduled tasks wisely**: Set appropriate intervals (daily vs hourly)
5. **Let auto-routing work**: The model auto-selector picks the most cost-effective model for each task
6. **Free chat after depletion**: Even with zero credits, you can still chat with Core for free