Skip to main content
BenchLM
Data

Scale Labs SWE Atlas – Codebase QnA (SWE Atlas – Codebase QnA)

We show this table for reference; we do not rank on it.

A Scale Labs public leaderboard mirrored as display-only reference data. It does not affect BenchLM rankings.

Scale score on SWE Atlas – Codebase QnA — November 12, 2026 snapshot

We mirror the published scale score view for SWE Atlas – Codebase QnA. Opus 5 (Claude Code) xHigh\n leads the public snapshot at 63.2%, followed by Fable-5.1 (Claude Code) xHigh* (60.0%) and GPT 6 Astra (Codex) xHigh* (59.1%). We do not use these results to rank models overall.

1

Opus 5 (Claude Code) xHigh\n

Anthropic

63.2%
2

Fable-5.1 (Claude Code) xHigh*

Anthropic

60.0%
3

GPT 6 Astra (Codex) xHigh*

OpenAI

59.1%

24 modelsCodingCurrentDisplay onlyUpdated November 12, 2026 snapshot

Scale score table (24 models)

Scale score results for SWE Atlas – Codebase QnA
RankModel / configurationScoreParameters (B)Open / closed
1Opus 5 (Claude Code) xHigh\nAnthropic
63.2%
Not reportedUnknown
2Fable-5.1 (Claude Code) xHigh*Anthropic
60.0%
Not reportedUnknown
3GPT 6 Astra (Codex) xHigh*OpenAI
59.1%
Not reportedUnknown
4Opus 4.8 (Claude Code) xhighAnthropic
57.3%
Not reportedUnknown
5GLM 5.2 (Mini-SWE-Agent)Z.AI
48.1%
Not reportedUnknown
6Gemini 3.8 Flash (Mini-SWE-Agent)Google
47.0%
Not reportedUnknown
7GPT-5.6-Sol (Codex) xHigh*OpenAI
46%
Not reportedUnknown
8GPT 5.5 (Codex) xHighOpenAI
45.4%
Not reportedUnknown
9Muse Spark 1.1 (Mini-SWE-Agent) xHighMeta
42.2%
Not reportedUnknown
10GPT 5.4 (Codex) xHigh\nOpenAI
40.8%
Not reportedUnknown
11Opus 4.7 (Claude Code)Anthropic
40.3%
Not reportedUnknown
12Fable-5 (Claude Code) xHigh*\nAnthropic
39%
Not reportedUnknown
13Gpt 5.4 xHigh (Mini-SWE-Agent)OpenAI
36.3%
Not reportedUnknown
14Opus 4.6 (Claude Code)Anthropic
33.3%
Not reportedUnknown
15GPT 5.3 (Codex) xHighOpenAI
32.6%
Not reportedUnknown
16Sonnet 4.6 (Claude Code)Anthropic
31.2%
Not reportedUnknown
17Opus 4.6 (Mini-SWE-Agent)Anthropic
30%
Not reportedUnknown
18DeepSeek V4 Pro (Mini-SWE-Agent)Deepseek
27.1%
Not reportedUnknown
19Muse SparkMeta
24.2%
Not reportedClosed
20Glm 5 (Mini-SWE-Agent)Z.AI
20.5%
Not reportedUnknown
21Gemini 3.1 Pro (Mini-SWE-Agent)Google
13.5%
Not reportedUnknown
22Kimi K2.5 (Mini-SWE-Agent)Moonshot AI
13.1%
Not reportedUnknown
23Minimax M2.5 (Mini-SWE-Agent)MiniMax
10.3%
Not reportedUnknown
24Gemini 3 Flash (Mini-SWE-Agent)Google
8.2%
Not reportedUnknown

How to read this leaderboard

Compare the published configurations as complete evaluation systems. The source can combine a base model, agent scaffold, tools, budget, and inference setting in each result.

Operator receipt: 24 sourced rows are currently displayable on this page; the leading published row is Opus 5 (Claude Code) xHigh\n at 63.2%.

Honest limit: This Scale table is display-only context, not benchmark provenance or a weighted model-only comparison.

SWE Atlas – Codebase QnA on Scale Labs

How BenchLM shows SWE Atlas – Codebase QnA

BenchLM mirrors 24 published rows from Scale Labs’ public SWE Atlas – Codebase QnA leaderboard, captured on November 12, 2026 snapshot.

The table is display only. It is useful context for a published agent or model configuration, but it does not enter BenchLM’s overall or category rankings.

Snapshot

24 published rowsScale Labs sourceDisplay only
SWE Atlas – Codebase QnA on Scale Labs

The published SWE Atlas – Codebase QnA snapshot places Opus 5 (Claude Code) xHigh\n first at 63.2%. The third row is 4.0 points behind. The broader top-10 range is 22.4 points, so the table still separates the published systems.

24 models have been evaluated on SWE Atlas – Codebase QnA. The benchmark falls in the Coding category. We keep external benchmark mirrors separate from the weighted global scoring system, so these results remain source-specific evidence. SWE Atlas – Codebase QnA is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About SWE Atlas – Codebase QnA

Year

2026

Tasks

24 published rows

Format

Published Scale leaderboard score

Difficulty

External agent and model evaluation

BenchLM mirrors 24 published rows from the SWE Atlas – Codebase QnA public table captured on November 12, 2026 snapshot.

Scale Labs leaderboardPublic benchmark source

Freshness and provenance

Version

SWE Atlas – Codebase QnA 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Questions

What does SWE Atlas – Codebase QnA measure?

A Scale Labs public leaderboard mirrored as display-only reference data. It does not affect BenchLM rankings.

Which model leads the published SWE Atlas – Codebase QnA snapshot?

Opus 5 (Claude Code) xHigh\n currently leads the published SWE Atlas – Codebase QnA snapshot with 63.2% scale score. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on SWE Atlas – Codebase QnA?

The November 12, 2026 snapshot snapshot contains 24 AI models.

Last updated: November 12, 2026 snapshot · mirrored from the public benchmark leaderboard

Know when it’s worth switching models

The model to choose, the cheaper alternative, and the release we would wait on.

Read a sample issue

Join 7,000+ subscribers.

One email each week. Unsubscribe anytime.