Measured · StudyNo views yet

e2e reference lens 1790522803022

GSM8K · 3 tasks · 1 harness · 1 model

GSM8K augmentation delivered the most efficient performance in this configuration.

Ranked shootout, no baseline arm: cells ranked by the composite score — the benchmark's own graders, then efficiency — claude-consumer / claude-sonnet-5 leads.

Abstract

This single-arm shootout evaluated three tasks with GSM8K augmentation across one harness/model cell. The treatment achieved a median cost of $0.0008 (mean $0.0009), consumed 159 median tokens (mean 138), and completed in a median 1.5s (mean 1.4s). Claude-consumer / claude-sonnet-5 ranked as the leader across cells. The benchmark was decided on efficiency grounds.

The result

Color by

claude-consumer

  1. 1claude-consumer / claude-sonnet-5100.0best

Overall score per cell, 0–100 points (not a pass rate): 75% benchmark pass rate (the benchmark's own graders carry the outcome share — no judge grades or evals on this board) + 25% efficiency (cost · tokens · duration, vs the board's best arm). Absent components renormalize. Best at the top — the board's best setup is tagged.

Leaderboard

Cells ranked by the composite score — outcome (the lenses that graded) × efficiency — absolute values, no baseline arm. Tokens, cost and duration are per-cell medians, the efficiency the score folds.

#HarnessModelScorePass rateQualityEvalsGoalTokensCostDurationRuns
1claude-consumerclaude-sonnet-5100.0100%−−100%159$0.00081.5s3/3

Pass rate

100%over 3 graded runs
CellPass rateTasksRunsRan asMatcher
claude-consumer/claude-sonnet-5100%33as publishednumeric

Each run's answer was extracted from its response and matched against the benchmark's own key — no model in the loop. Two-stage: the mean over repeats per task, then the mean over tasks, with the interval taken over tasks.

Quality × efficiency clusters — normalized across the study

Every graded run, standardized across the study — each task ran once, so there is no field per task to compare against: → right = fewer tokens than the study's average run, ↑ up = higher quality index (85% the benchmark's own pass/fail + 15% outcome) than average. Every run scored alike on quality here, so the vertical axis separates nothing — read the map left to right. The field pools every harness and model, so a setup's left–right position largely reflects its own token habits. Dots are runs; each harness logo is a harness × model × arm centroid — hover it for the model and averages. Up-right wins.

Task
3 of 3 runs match
Show
Color by

claude-consumer

better · cheaperworse · pricier
← pricier than the fieldefficiency (σ, per task)cheaper than the field →

Every metric, per harness × model

Benchmark pass rate

  1. claude-consumer / claude-sonnet-5100%

Task completion

  1. claude-consumer / claude-sonnet-5100%

Cost per run

  1. claude-consumer / claude-sonnet-5$0.0008

Tokens per run

  1. claude-consumer / claude-sonnet-5159

Duration per run

  1. claude-consumer / claude-sonnet-51.5s

Distributions

Every completed run is one dot — the spread the averages hide. Click a dot to replay that run's journey.

Cost per run

Your setup
consumer/claude-sonnet-5
0$0.0012

Token composition — average per run

InputOutputCache readCache write
consumer/claude-sonnet-5 · treat
138

Statistics

MetricArmnMeanMedian (pooled)MinMaxStd dev
Costtreat3/3$0.0009$0.0008$0.0006$0.0012$0.0003
Durationtreat3/31.4s1.5s1.3s1.6s113ms
Total tokenstreat3/31381599615936
Output tokenstreat3/34031256320
Turnstreat3/311110

n = runs carrying the fact / completed runs in the arm; every statistic runs over present facts only — a missing fact is never counted as 0. Std dev is the sample form (n−1), withheld below n = 2. Every figure here POOLS all runs. The typical-run figures in the abstract, the head-to-head decision and the per-setup panels take each task's median first and then the median across tasks, so a task that ran more often never outweighs one that ran once; the two medians can differ. A promise study's headline (“cut average tokens”) compares per-arm means.

What each setup did

Derived from each run's recorded tool calls — not from a model's description of the run — and aggregated per setup, so a behavior seen across several runs is stated once with its rate. Open a finding to see the runs behind it, each linked to its journey at the step where it happened.

Task
Pick a finding or a setup to list the runs behind it.
claude-consumer · claude-sonnet-5 · Your setup1 finding
  • 3 of 3 runs · 3 completed

Task text withheld

Task text withheld — this benchmark is guarded and its tasks are not republished here.

Tasks

Runs

Every run is inspectable — open one to replay the agent's journey step by step, with the analyst's read underneath.

Every run, filterable… of 3Show runs
Task
… of 3 runs match
Sort
HarnessModelTaskCompletedPassTokensCostDuration

Loading runs…

Methodology

What each metric means
Completed
The run finished and the harness returned a response. It does NOT mean the answer was correct — a completed run can score zero on quality.
Denominator: Terminal runs, excluding those killed by our own infrastructure.
Quality index
A composite used only in the quality x efficiency plot, built from whichever grader scored this study's runs — the judge's rubric, the pre-registered checks or the benchmark's own verdict — plus whether the run's own outcome was success. The map's caption names the grader and its weights.
Denominator: Graded runs — runs the study's grader scored.
Infra-excluded
A run killed by our own infrastructure. It is a missing measurement, never a loss for the arm, and is excluded from every rate denominator.
Denominator: Reported as a count beside every affected panel.
Single arm
Every task runs once per harness × model cell — a shootout with no baseline arm. The readout is absolute (quality, success, tokens, cost) and the cells are ranked into a leaderboard.
One run per task
Every task ran once per cell, in one phrasing. Run-to-run variation is therefore not measured: a single task's difference can be one lucky or unlucky attempt. Every row of the reference table treats the TASKS as the unit: its intervals resample the tasks and its p-values come from a paired test over them, so they describe how much the result depends on which tasks were drawn — not how it would change if the same tasks were run again. With few tasks no difference can reach significance (five tasks cannot go below p = 1/16).
Sample size
1 cell × 3 tasks × 1 arm = 3 runs.