AI Workflow Benchmark

Compare recorded tool and workflow evidence

{{ total_results }} Total Runs

{{ task_results|length }} Tasks

{{ tools|length }} Experiment Cohorts

{% if tools %}

Swipe or scroll sideways to view all metrics.

{% for tool in tools %} {% endfor %}

Rank	Tool	Model	Composite	Success %	Avg Score %	Avg Time (s)	Avg Cost ($)	Avg Iterations
{{ tool.rank if tool.rank is not none else "n/a" }}	{{ tool.tool }} Cohort {{ tool.identity_hash }}{% if not tool.comparison_eligible %} Not comparison eligible: {{ tool.ineligibility_reasons\|length }} unmet requirements{% endif %}	{{ tool.model }}	{% if tool.composite_score is not none %}0{% else %}n/a{% endif %}	0%	0%	{{ tool.avg_time }}	{% if tool.avg_cost is not none %}${{ tool.avg_cost }}{% else %}n/a{% endif %}	{{ tool.avg_iterations }}

{% if comparison_eligible_count %}

Metric Comparison

{% endif %} {% if task_results %}

{% for task_id, tool_results in task_results.items() %}

{% for tool_name, result in tool_results.items() %} {% endfor %}

Tool	Success	Score	Time (s)	Cost ($)
{{ tool_name }}	{% if result.outcome.success %}PASS{% else %}FAIL{% endif %}	{{ result.outcome.partial_credit_score }}/{{ result.outcome.partial_credit_max }}	{{ result.metrics.wall_clock_seconds }}	${{ result.cost.estimated_cost_usd }}

{% endfor %}

{% endif %} {% else %}

Run benchmarks first:

awb run claude-code-vanilla -t BF-001
awb leaderboard

{% endif %}