Skip to main content
BenchLM
Data

Scale Labs SEAL Agentic Tool Use (Enterprise) (SEAL Agentic Tool Use (Enterprise))

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

ToolComp-Enterprise: the subset of Scale’s 485-prompt tool-use benchmark that needs models to chain calls across 11 enterprise tools. BenchLM shows it as display-only reference data, excluded from rankings. Scale has not added a row to this board since March 2025.

Accuracy (%) on SEAL Agentic Tool Use (Enterprise) — November 12, 2026 snapshot

We mirror the published accuracy (%) view for SEAL Agentic Tool Use (Enterprise). o1 (December 2024) leads the public snapshot at 70.1%, followed by Gemini 2.5 Pro Experimental (March 2025) (68.8%) and o1 Pro (March 2025) (67.0%). We do not use these results to rank models overall.

1

o1 (December 2024)

OpenAI

70.1%
2

Gemini 2.5 Pro Experimental (March 2025)

Google

68.8%
3

o1 Pro (March 2025)

OpenAI

67.0%

35 modelsAgenticCurrentDisplay onlyUpdated November 12, 2026 snapshot

Accuracy (%) table (35 models)

Accuracy (%) results for SEAL Agentic Tool Use (Enterprise)
RankModel / configurationScoreParameters (B)Open / closed
1o1 (December 2024)OpenAI
70.1%
Not reportedUnknown
2Gemini 2.5 Pro Experimental (March 2025)Google
68.8%
Not reportedUnknown
3o1 Pro (March 2025)OpenAI
67.0%
Not reportedUnknown
4o1-previewOpenAI
66.4%
Not reportedClosed
5Claude 3.7 Sonnet Thinking (February 2025)Anthropic
65.3%
Not reportedUnknown
6DeepSeek-R1DeepSeek
65.3%
Not reportedOpen
7o3-mini (high)OpenAI
65.3%
Not reportedUnknown
8Claude 3.7 Sonnet (February 2025)Anthropic
64.9%
Not reportedUnknown
9o3-mini (medium)OpenAI
64.9%
Not reportedUnknown
10GPT-4o (May 2024)OpenAI
64.6%
Not reportedUnknown
11DeepSeek V3 (March 2025)Deepseek
64.2%
Not reportedUnknown
12GPT-4.5 Preview (February 2025)OpenAI
63.8%
Not reportedUnknown
13Gemini 2.0 Flash Thinking Experimental (January 2025)Google
63.2%
Not reportedUnknown
14DeepSeek-V3 (December 2024)Deepseek
62.5%
Not reportedUnknown
15Gemini 2.0 Pro Experimental (February 2025)Google
61.5%
Not reportedUnknown
16GPT-4 Turbo PreviewOpenAI
60.8%
Not reportedUnknown
17Gemini 1.5 Pro (August 27, 2024)Google
60.3%
Not reportedUnknown
18Gemini 2.0 Flash Experimental (December 2024)Google
60.1%
Not reportedUnknown
19GPT-4o (August 2024)OpenAI
59.9%
Not reportedUnknown
20Claude 3.5 Sonnet (June 2024)Anthropic
59.4%
Not reportedUnknown
21Gemini 2.0 FlashGoogle
55.6%
Not reportedClosed
22Claude 3 SonnetAnthropic
54.2%
Not reportedUnknown
23Gemini 2.0 Flash Lite Preview (February 2025)Google
53.8%
Not reportedUnknown
24Claude 3 OpusAnthropic
52.8%
Not reportedClosed
25GPT-4o miniOpenAI
51.7%
Not reportedClosed
26GPT-4OpenAI
51.4%
Not reportedUnknown
27Llama 3.1 405B InstructMeta
50.4%
Not reportedUnknown
28Mistral Large 2Mistral
50.4%
Not reportedClosed
29Nova ProAmazon
47.0%
Not reportedClosed
30Gemini 1.5 Pro (May 2024)Google
40.4%
Not reportedUnknown
31Llama 3.1 70B InstructMeta
37.2%
Not reportedUnknown
32Nova LIteAmazon
36.5%
Not reportedUnknown
33Command R+Cohere
30.2%
Not reportedUnknown
34Nova MircoAmazon
28.9%
Not reportedUnknown
35Llama 3.1 8B InstructMeta
17.4%
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: 35 sourced rows are currently displayable on this page; the leading published row is o1 (December 2024) at 70.1%.

Honest limit: This Scale table is display-only context, not benchmark provenance or a weighted model-only comparison. Scale has not added a row to this board since March 2025.

SEAL Agentic Tool Use (Enterprise) on Scale Labs

How BenchLM shows SEAL Agentic Tool Use (Enterprise)

BenchLM mirrors 35 published rows from Scale Labs’ public SEAL Agentic Tool Use (Enterprise) 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 has not added a row to this board since March 2025.

Snapshot

35 published rowsScale Labs sourceNo new rowsDisplay only
SEAL Agentic Tool Use (Enterprise) on Scale Labs

The published SEAL Agentic Tool Use (Enterprise) snapshot places o1 (December 2024) first at 70.1%. The third row is 3.1 points behind. The broader top-10 range is 5.6 points, so many of the published results sit in a relatively narrow band.

35 models have been evaluated on SEAL Agentic Tool Use (Enterprise). The benchmark falls in the Agentic category. We keep external benchmark mirrors separate from the weighted global scoring system, so these results remain source-specific evidence. SEAL Agentic Tool Use (Enterprise) is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About SEAL Agentic Tool Use (Enterprise)

Year

2026

Tasks

ToolComp prompts, Enterprise subset

Format

Accuracy (%)

Difficulty

External agent and model evaluation

ToolComp prompts require several dependent tool calls, with golden answer chains and step-level labels. The Enterprise subset uses 11 tools. The score is final-answer accuracy in percent. BenchLM mirrors 35 published rows captured on November 12, 2026 snapshot.

Scale Labs leaderboardPublic benchmark source

Freshness and provenance

Version

SEAL Agentic Tool Use (Enterprise) 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 Agentic Tool Use (Enterprise) measure?

ToolComp-Enterprise: the subset of Scale’s 485-prompt tool-use benchmark that needs models to chain calls across 11 enterprise tools. BenchLM shows it as display-only reference data, excluded from rankings. Scale has not added a row to this board since March 2025.

Which model leads the published SEAL Agentic Tool Use (Enterprise) snapshot?

o1 (December 2024) currently leads the published SEAL Agentic Tool Use (Enterprise) snapshot with 70.1% accuracy (%). BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on SEAL Agentic Tool Use (Enterprise)?

The November 12, 2026 snapshot snapshot contains 35 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.