← christiant.io

Coding Agent Explorer

Compare coding-agent configurations across the Artificial Analysis Coding Agent Index v1.4, current cost per task, and token utilization.

Configurations shown
0
of 66 in dataset
Pareto knee
--
--
Lowest cost per task
--
--
Median token use
--
Average total tokens per task
1970
$0.05$12
Axis scaling Choose linear or logarithmic scaling independently for each chart dimension.
Cost / task
Coding Agent Index
Tokens / task
Display

Current source-cost view. Cost / Task comes directly from Artificial Analysis v1.4. There is no OpenAI-specific adjustment layer and no historical price toggle. These source values drive the cost filter, x-axis, Pareto frontier, zone thresholds, metrics, and detail panel.

3D scatter cube

Hover a point for model details. Drag to rotate, scroll to zoom, or click to pin details.

No configurations match the current filters.

Select a configuration

Coding Agent IndexHover or click a point
Cost per task--
Tokens per task--

Model company legend

Source: Artificial Analysis Coding Agent Index v1.4, retrieved September 3, 2026. The Index equally weights DeepSWE, Terminal-Bench v2.1, and SWE-Atlas-QnA. Cost and token utilization are Artificial Analysis per-task averages using current pay-per-token pricing and provider cache treatment where applicable. No local price multipliers are applied. Colors identify model company; harness is shown in hover and selection details.