`RunResult` missing `.usage` attribute in latest OpenAI Agents SDK

[LatVAlY]

Description

I'm using the latest openai-agents Python SDK and running an agent successfully with Runner.run(...).
However, the returned RunResult object does not have a .usage attribute, even though the documentation suggests it should.

Code Snippet

base_result = await Runner.run(
    self.agent,
    input=instruction
    context=context,
)

print(base_result.final_output)   # works fine
print(base_result.usage)          # raises AttributeError

Error

AttributeError: 'RunResult' object has no attribute 'usage'

Environment

• openai-agents version: 0.2.6
• Python: 3.11
• OS: (e.g., macOS 14.5 / Ubuntu 22.04)
• Agent Type: Custom agent built with Agent and run via Runner.run(...)

Expected behavior

The RunResult object should contain .usage with input_tokens, output_tokens, and total_tokens, as per the documentation/examples.

Additional context

This is blocking me from tracking token consumption per run.
If .usage is not supported in Runner.run, please clarify the intended way to retrieve token usage for agent runs.

[dast-draftwise]

• 1 need help with that as well.
  The documentation is horrible. https://openai.github.io/openai-agents-python/ref/usage/
  Please add some examples!

[khomiakovkevin]

@LatVAlY @dast-draftwise - I think the correct way to retrieve the usage is shown here. Please tell me if this is helpful.

[khomiakovkevin]

I think this issue can be closed