Original engineering problems test coding agents on substantial changes in existing repositories.
88 model results
One row per model and compatible score measure. Mean = equal-weight arithmetic average of imported reports. Distinct configurations reported for the same model in a source table remain separate. Reports can overlap; the range is not a confidence interval. Unbundle to inspect each observation.
Published observations span 0.3 percentage points. Configuration and source differences contribute to this spread; it is not measured run-to-run uncertainty.
#9 of 82 in this displayed comparison; 5.15 pp below the leader.
Published observations span 0.3 percentage points. Configuration and source differences contribute to this spread; it is not measured run-to-run uncertainty.
#17 of 82 in this displayed comparison; 6.15 pp below the leader.
Published observations span 1.01 percentage points. Configuration and source differences contribute to this spread; it is not measured run-to-run uncertainty.
#30 of 82 in this displayed comparison; 10.99 pp below the leader.
Published observations span 2.06 percentage points. Configuration and source differences contribute to this spread; it is not measured run-to-run uncertainty.
#31 of 82 in this displayed comparison; 11.07 pp below the leader.
Values retain the model card’s original formatting and score units. HLE is without tools; the final three visual evaluations use tools. The dagger marks source qualifications: inspect the original report before treating those results as matched.
A transcription of the October model card, including its external leaderboard snapshots. These are not uniformly matched runs. Effort, tools, hardware, frame budgets and task subsets differ. GraphWalks reports F1; OSWorld uses the offline subset and partial credit. Harvey refers to the archived Vals v1 protocol, not the remediated v1.1 or AA variant. Consult the linked methodology.
Launch-reported maxima across reasoning efforts, not a uniformly matched effort comparison. Research/API settings can differ from ChatGPT. Source footnotes disclose altered harnesses, model revisions and reproduced competitor results. A dagger marks cells with additional source qualifications. Preserve each row’s named subset and tools.
Source-reported model settings are max unless a cell says otherwise. The original page reports numbers without a uniform percent symbol; original formatting is retained. OSWorld includes two scoring rules in each cell: leaderboard ordering uses partial credit only. The internal composite and AA index use separate scales; their scores cannot be averaged.
Headline measurements transcribed from the official launch text. The report names OSWorld V2 without fully specifying its subset and scoring details in the prose. A single-model cohort is evidence of a reported run, not a competitive ranking.
Epoch AI public data snapshot, CC BY 4.0. Externally collected result; original evaluation provider is linked. CSV: deepswe_external.csv; score column: Pass@1; fraction scale: 1. Model version is retained in configuration. Original source authors retain credit and any additional licensing terms.
DeepSWE v1.1 · Kimi K3.1 Pro · complete release comparison · 2026-10-27
Published model comparison, 2026. Kimi K3.1 Pro · complete release comparison. Exact values from the publisher’s embedded full comparison table. Maximum effort, fallback settings and dated model versions remain in model labels. Blank cells stay missing. Source footnotes specify mixed harnesses and externally quoted comparisons; this is a launch-reported cohort rather than a new independent evaluation. Benchmarks · full embedded comparison table · row 7; source precision and any reported uncertainty retained. These are source-reported observations, not evaluations run by RLTasks.
TRACE THE SCORE · Per-model notes use published breakdowns where available.
Score vs. release date
Epoch AI · 69 observations
DeepSWE · Pass@1 · % · higher is better
Dates are model release dates recorded by Epoch AI, not evaluation dates. Each point retains its published score and configuration. Separate releases are not a matched longitudinal experiment. Missing dates are omitted. Explore all reported results ↓
The dashed line traces the best observed score through the model release dates. It describes this selection of reports, with their original configurations. It does not estimate unmeasured models or costs.
View plotted values and sources (69)
DeepSWE v1.1 · DeepSWE · Pass@1 · %. Published measurements in their original score units.
These model reports include this benchmark. Their harnesses, task subsets, scoring policies, and resource budgets can differ; follow each report’s evaluation notes before joining the numbers.
Original engineering problems test coding agents on substantial changes in existing repositories.
How to read the result
Inspect the execution harness and committed-code verifier, not just the model name. An agent can implement a sound change that an imperfect grader fails to recognize.
Limits and comparison traps
Official and lab-run results can use different scaffolds. Independent audits have reported grading defects, so small score gaps need particular care.
Epoch data and evaluation setting
Epoch AI publishes this benchmark in its capability collection. Imported rows retain the published score column and model configuration. Original authors and evaluation providers remain credited in the result references.
Epoch attribution and score coverage
Data is attributed to Epoch AI under CC BY 4.0, with original evaluation providers credited in the linked source snapshots. Scores are imported only where a published mapping identifies the field and its scale; missing or incompatible records are omitted rather than assigned zero.
Published evaluation
Exact values from the publisher’s embedded full comparison table. Maximum effort, fallback settings and dated model versions remain in model labels. Blank cells stay missing. Source footnotes specify mixed harnesses and externally quoted comparisons; this is a launch-reported cohort rather than a new independent evaluation.
Where this fits in an evaluation suite
The agent wrapper is part of the system. Pin tools, dependencies, timeouts, tests, and the repository revision. Compare contest coding separately from repository maintenance.
Start with a capability hypothesis, then choose a second test with a different failure mode. Keep the checkpoint, reasoning effort, agent tools, sampling policy, and scoring procedure fixed when measuring a training intervention. Record failures and missing outputs in the denominator.
A reproducible comparison
Exact values from the publisher’s embedded full comparison table. Maximum effort, fallback settings and dated model versions remain in model labels. Blank cells stay missing. Source footnotes specify mixed harnesses and externally quoted comparisons; this is a launch-reported cohort rather than a new independent evaluation.
Publish the exact run configuration, item coverage, score units, and available uncertainty. If a source supplies only an aggregate or a single run, preserve that limit. A new model name or a rounded score alone does not establish a capability improvement.