Claim · coding

GPT-5.6 Lunaon DeepSWE

1.6 to 67.2 the range across every scoreable measurement we hold
65.6 points between the highest and the lowest
44.3 the number the board uses: the lower median, a real row with its own provenance
5 declared configurations, from 1 runner
5 / 5 rows whose cited page still shows this figure, refetched and checked by us

All 5 scoreable measurements here come from ONE runner, so this is not a reconciliation and is not presented as one. What it shows is how far a single runner's number moves when the model is run differently.

Every measurement

5 rows. Best first. the board takes the LOWER MEDIAN of the scoreable measurements, which is a real row with its own interval and provenance rather than an interpolated average. a vendor self-report is shown beside the independent numbers and is never one of them; it cannot move a rank.

ScoreProvenanceConfigurationWho ran itWhere to check it
67.2 / 100
as published: 67.2%
INDEPENDENT
counts toward the ranking
gpt-5.6-luna
harness mini-swe-agent
effort max
deepswe.datacurve.ai
via Epoch AI
CC-BY-4.0
checked: this figure appears in the page’s row for this model2026-08-30model:"gpt-5-6-luna",harness:"mini-swe-agent",reasoning_effort:"max",config:"mini_swe_agent_gpt_5_6_luna_max",source:"deep-swe",pass_rate:0.671875,pass_at_1:0.6
the leaderboard this came from
the leaderboard for this benchmark, not a link to this row
56.9 / 100
as published: 56.9%
INDEPENDENT
counts toward the ranking
gpt-5.6-luna
harness mini-swe-agent
effort xhigh
deepswe.datacurve.ai
via Epoch AI
CC-BY-4.0
checked: this figure appears in the page’s row for this model2026-08-30model:"gpt-5-6-luna",harness:"mini-swe-agent",reasoning_effort:"xhigh",config:"mini_swe_agent_gpt_5_6_luna_xhigh",source:"deep-swe",pass_rate:0.5685840707964602
the leaderboard this came from
the leaderboard for this benchmark, not a link to this row
44.3 / 100
as published: 44.2%
INDEPENDENT
counts toward the ranking
gpt-5.6-luna
harness mini-swe-agent
effort high
deepswe.datacurve.ai
via Epoch AI
CC-BY-4.0
checked: this figure appears in the page’s row for this model2026-08-30model:"gpt-5-6-luna",harness:"mini-swe-agent",reasoning_effort:"high",config:"mini_swe_agent_gpt_5_6_luna_high",source:"deep-swe",pass_rate:0.4424778761061947,p
the leaderboard this came from
the leaderboard for this benchmark, not a link to this row
11.3 / 100
as published: 11.3%
INDEPENDENT
counts toward the ranking
gpt-5.6-luna
harness mini-swe-agent
effort medium
deepswe.datacurve.ai
via Epoch AI
CC-BY-4.0
checked: this figure appears in the page’s row for this model2026-08-30model:"gpt-5-6-luna",harness:"mini-swe-agent",reasoning_effort:"medium",config:"mini_swe_agent_gpt_5_6_luna_medium",source:"deep-swe",pass_rate:0.11283185840707
the leaderboard this came from
the leaderboard for this benchmark, not a link to this row
1.6 / 100
as published: 1.5%
INDEPENDENT
counts toward the ranking
gpt-5.6-luna
harness mini-swe-agent
effort low
deepswe.datacurve.ai
via Epoch AI
CC-BY-4.0
checked: this figure appears in the page’s row for this model2026-08-30model:"gpt-5-6-luna",harness:"mini-swe-agent",reasoning_effort:"low",config:"mini_swe_agent_gpt_5_6_luna_low",source:"deep-swe",pass_rate:0.015486725663716814,p
the leaderboard this came from
the leaderboard for this benchmark, not a link to this row
← The Forge Rankings

Snapshot 2026-08-30-62e4e85f043b, manifest dc46131315046f1e, as of 2026-08-30. where only one party has measured a model on a benchmark, this page says so rather than presenting a single number as a reconciliation.