Atlas

Benchmarks

Ramp SWE-Bench

Ramp SWE-Bench is a private, production-grounded coding benchmark built from real Ramp backend engineering tasks. Its canonical manifest contains 80 human-accepted tasks evaluated using mini-swe-agent, with pass rate reported as the percentage of tasks resolved. Published runs omitting task IDs 35 and 75 belong to the separate ramp-swe-bench-observed-78 cohort.

Models
0
Top score
—
Leader
—
Updated
—
Resolution
—
Saturation
—

No signal or difficulty estimate. No usable scores.

This is a benchmark-family page. Scores are recorded on the component benchmarks below rather than directly on the parent, because the components use different score scales and directions.

MetricModelsLeaderDirectionDate
Ramp SWE-Bench (78 published tasks)26Claude Fable 588.5higher is better—
Loading Atlas data…