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

# Usage

> See the tokens, estimated cost, sessions and cache hit rate of your AI Studio conversations over the last 7, 30 or 90 days

The **Usage** page shows how much of the model you used in AI Studio: tokens, model calls, estimated cost and cache hit rate, per day, per model and per conversation.

Open it from **Usage** in the AI Studio sidebar (route `/aistudio/usage`). It needs the `agent.read` permission.

![AI Studio Usage page with the Last 30 days range selected, six cards (Cost n/a, Sessions 13, Requests 365, Input tokens 22.1M with 19.2M read from cache, Output tokens 313.3K, Cache hit 87%), a Tokens per day bar chart, a By model table with one custom model, and a Top sessions list](https://mintcdn.com/oasmplatform/JcAER2Ib37AN8_8A/images/screens/ai-usage.webp?fit=max\&auto=format\&n=JcAER2Ib37AN8_8A\&q=85\&s=81450b96722114095d6f1798dc5027dd)

The Usage page over the last 30 days

## Whose usage is shown

The page shows **your own** usage in the **current workspace**:

* Every finished answer in your conversations, and every subagent run those conversations dispatched.
* Answers you stopped are not counted.
* Background calls, such as title generation and conversation summaries, are not counted.
* Usage belongs to its conversation: deleting a conversation removes its usage from this page.

Other members' usage is never shown, and neither is usage in other workspaces. To see what one member spends, that member opens their own Usage page.

## Choose a time range

Use the range menu at the top right: **Last 7 days**, **Last 30 days** (the default) or **Last 90 days**.

Days are calendar days in your browser's time zone, ending today. An answer counts on the day it started; a subagent run counts on the day it completed.

If nothing was used in the range, the page shows "No AI Studio usage in the last N days."

## Summary cards

| Card | Value | Line underneath |
| - | - | - |
| **Cost** | Estimated cost in US dollars (see [How cost is estimated](#how-cost-is-estimated)) | **Estimated from model prices**, or **Some models have no known price** |
| **Sessions** | Conversations with usage in the range | Number of answers and subagent runs, for example `21 answers · 9 subagent runs` |
| **Requests** | Model calls. An answer that calls tools makes one model call per step | **Model calls** |
| **Input tokens** | Tokens sent to the model, including those served from cache | How many were read from cache |
| **Output tokens** | Tokens the model generated, including reasoning tokens | Input and output tokens in total |
| **Cache hit** | Share of input tokens served from the provider's cache | **Of input tokens, served from cache** |

Token counts are shortened, for example `22.1M` or `313.3K`.

## Tokens per day

A bar chart of the total tokens (input plus output) for each day of the range. Hover a bar to see that day's **Tokens**, **Input**, **Output**, **Cache hit**, **Requests** and **Cost**.

## By model

One row per model and provider, sorted by total tokens, heaviest first.

| Column | Description |
| - | - |
| **Model** | The model ID, followed by the provider (for example `custom`). The bar under it compares the model's total tokens with the heaviest model |
| **Requests** | Model calls |
| **Input** | Input tokens |
| **Output** | Output tokens |
| **Cache hit** | Share of input tokens served from cache |
| **Cost** | Estimated cost |

Subagent runs are counted under the model they ran on, so a separate [Subagent Model](https://docs.oasm.io/ai-studio/settings/models.md#subagent-model) shows up as its own row.

## Top sessions

Up to ten conversations with the most tokens in the range. Each row shows:

* The conversation title, and when it was last active.
* Input and output tokens, for example `8.9M in · 106.4K out`.
* Cache hit and cost, for example `91% cache · n/a`.

Click a row to open the conversation.

The footer note reads "Your finished answers and subagent runs in this workspace. Stopped answers are not counted."

## How cost is estimated

When an answer or a subagent run finishes, OASM computes its cost from the model's published prices per million tokens. The prices come from the [models.dev](https://models.dev) catalog snapshot that ships with your OASM version:

* Input tokens not served from cache, at the input price.
* Tokens read from cache, at the cache-read price (or the input price if the model has none).
* Tokens written to cache, at the cache-write price (or the input price if the model has none).
* Output tokens, including reasoning tokens, at the output price.

A cost is shown as:

| Display | Meaning |
| - | - |
| `$1.20` | Every call in it had a known price |
| `≥ $1.20` | Some calls had no known price, so the real cost is at least this |
| `n/a` | No call in it had a known price |
| `$0` | All calls had a known price of zero |

### Why cost shows n/a

A call has no known price when:

* It ran on the **Custom provider (OpenAI-compatible)**. OASM never prices self-hosted or custom endpoints.
* Its model is not in the bundled catalog for that provider, for example a model newer than your OASM version or a gateway's own routing model.

The cost is an estimate. Your provider's invoice is the source of truth: discounts, price changes after your OASM release, and provider fees are not reflected.

## Cache hit

Providers can cache the beginning of a prompt that repeats between calls, such as the system prompt, tool definitions and earlier turns of the conversation. **Cache hit** is the share of input tokens the provider served from that cache:

```
cache hit = tokens read from cache / input tokens
```

A high cache hit lowers cost when the model has a cheaper cache-read price, which the cost estimate uses. Agent runs with many steps tend to reach a high rate, because each step resends the conversation so far. The value is colored green from 90%, light green from 70%, amber from 40% and red below that, and shows `—` when there were no input tokens.

A low cache hit can mean that the provider or model does not report cached tokens, or does not cache prompts at all.

## Related

> [card] [Models](https://docs.oasm.io/ai-studio/settings/models.md)
>
> Choose a cheaper subagent model or change the default model

> [card] [Side panel](https://docs.oasm.io/ai-studio/side-panel.md)
>
> Turn stats, context usage and cost of a single conversation

> [card] [Connect a model](https://docs.oasm.io/ai-studio/connect-a-model.md)
>
> Connect providers and pick models

> [card] [AI Studio overview](https://docs.oasm.io/ai-studio/overview.md)
>
> What AI Studio is and how an agent task flows


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.