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

# Replay Examples

> Examples for inspecting past runs and comparing outputs.

## 01 — Basic Replay

Demonstrates inspecting a completed run and comparing two runs side-by-side.

```python theme={null}
from definable.agent import Agent

agent = Agent(model=model)

# Run and inspect
output = agent.run("Summarize this document.")
replay = agent.replay(run_output=output)

print(f"Model: {replay.model}")
print(f"Status: {replay.status}")
print(f"Tokens: {replay.tokens.total_tokens}")
print(f"Cost: ${replay.cost:.4f}")
print(f"Tool calls: {len(replay.tool_calls)}")
print(f"Content: {replay.content}")

# Compare two runs
output2 = agent.run("Summarize this document.")
diff = agent.compare(output, output2)

print(f"Token diff: {diff.token_diff}")
print(f"Cost diff: ${diff.cost_diff:.4f}")
print(f"Content changed: {diff.content_diff is not None}")
print(f"Tools added: {len(diff.tool_calls_diff.added)}")
print(f"Tools removed: {len(diff.tool_calls_diff.removed)}")
```

<Card title="Full source" icon="github" href="https://github.com/definable-ai/definable/blob/main/definable/examples/replay/01_basic_replay.py">
  `definable/examples/replay/01_basic_replay.py`
</Card>