Skip to main content
BenchLM
Data

Scale Labs SEAL Coding (SEAL Coding)

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

Scale’s SEAL coding board: expert reviewers compared model answers to 1,000 prompts across programming languages, debugging, optimization, and documentation tasks. BenchLM shows it as display-only reference data, excluded from rankings. Scale marked this board deprecated as of March 2025 and no longer adds models.

Rating on SEAL Coding — November 12, 2026 snapshot

We mirror the published rating view for SEAL Coding. o1-mini leads the public snapshot at 1237, followed by o3-mini (1137) and GPT-4o (November 2024) (1132). We do not use these results to rank models overall.

1Closed

o1-mini

OpenAI

1237
Context 128K
2Closed

o3-mini

OpenAI

1137
Overall 38.5Context 200K
3

GPT-4o (November 2024)

OpenAI

1132

28 modelsCodingCurrentDisplay onlyUpdated November 12, 2026 snapshot

Rating table (28 models)

Rating results for SEAL Coding
RankModel / configurationScoreParameters (B)Open / closed
1o1-miniOpenAI
1237
Not reportedClosed
2o3-miniOpenAI
1137
Not reportedClosed
3GPT-4o (November 2024)OpenAI
1132
Not reportedUnknown
4o1-previewOpenAI
1123
Not reportedClosed
5Gemini 2.0 Flash Experimental (December 2024)Google
1111
Not reportedUnknown
6Gemini 2.0 Pro (December 2024)Google
1109
Not reportedUnknown
7Gemini 2.0 Flash Thinking (January 2025)Google
1108
Not reportedUnknown
8DeepSeek-R1DeepSeek
1100
Not reportedOpen
9o1 (December 2024)OpenAI
1083
Not reportedUnknown
10Claude 3.5 Sonnet (June 2024)Anthropic
1079
Not reportedUnknown
11GPT-4o (August 2024)OpenAI
1045
Not reportedUnknown
12GPT-4o (May 2024)OpenAI
1036
Not reportedUnknown
13GPT-4 Turbo PreviewOpenAI
1034
Not reportedUnknown
14Mistral Large 2Mistral
1029
Not reportedClosed
15Llama 3.1 405B InstructMeta
1022
Not reportedUnknown
16Gemini 1.5 Pro (August 27, 2024)Google
1007
Not reportedUnknown
17Gemini 1.5 Pro (May 2024)Google
994
Not reportedUnknown
18GPT-4 (November 2024)OpenAI
992
Not reportedUnknown
19DeepSeek V3DeepSeek
985
Not reportedOpen
20Llama 3.2 90B Vision InstructMeta
984
Not reportedUnknown
21Claude 3 OpusAnthropic
959
Not reportedClosed
22Gemini 1.5 FlashGoogle
943
Not reportedClosed
23Gemini 1.5 Pro (April 2024)Google
891
Not reportedUnknown
24Claude 3 SonnetAnthropic
879
Not reportedUnknown
25Llama 3 70B InstructMeta
871
Not reportedUnknown
26Mistral LargeMistral
811
Not reportedUnknown
27Gemini 1.0 ProGoogle
685
Not reportedClosed
28CodeLlama 34B InstructMeta
598
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: 28 sourced rows are currently displayable on this page; the leading published row is o1-mini at 1237.

Honest limit: This Scale table is display-only context, not benchmark provenance or a weighted model-only comparison. Scale marked this board deprecated as of March 2025 and no longer adds models.

SEAL Coding on Scale Labs

How BenchLM shows SEAL Coding

BenchLM mirrors 28 published rows from Scale Labs’ public SEAL Coding 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. Scale marked this board deprecated as of March 2025 and no longer adds models.

Snapshot

28 published rowsScale Labs sourceDeprecated by Scale (March 2025)Display only
SEAL Coding on Scale Labs

The published SEAL Coding snapshot places o1-mini first at 1237. The third row is 105 score units behind. The broader top-10 range is 158 score units, so the table still separates the published systems.

28 models have been evaluated on SEAL Coding. 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. SEAL Coding is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About SEAL Coding

Year

2026

Tasks

1,000 coding prompts

Format

Elo-style rating

Difficulty

External agent and model evaluation

Reviewers checked accuracy, coherence, and relevance, with code execution tests on the answers. Scores are ratings on an Elo-style scale built from expert human comparisons, so only the gaps between rows carry meaning. BenchLM mirrors 28 published rows captured on November 12, 2026 snapshot.

Scale Labs leaderboardPublic benchmark source

Freshness and provenance

Version

SEAL Coding 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 SEAL Coding measure?

Scale’s SEAL coding board: expert reviewers compared model answers to 1,000 prompts across programming languages, debugging, optimization, and documentation tasks. BenchLM shows it as display-only reference data, excluded from rankings. Scale marked this board deprecated as of March 2025 and no longer adds models.

Which model leads the published SEAL Coding snapshot?

o1-mini currently leads the published SEAL Coding snapshot with 1237 rating. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on SEAL Coding?

The November 12, 2026 snapshot snapshot contains 28 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.