Skip to main content
benchlm.ai

JobBench

An occupational agent benchmark for professional workflows that workers say they most want delegated to AI.

Data verified

Benchmark score on JobBench — July 9, 2026

BenchLM mirrors the published score view for JobBench. Muse Spark 1.1 leads the public snapshot at 54.7% , followed by Claude Opus 4.7 (Adaptive) (45.9%) and GPT-5.5 (42.7%). BenchLM does not use these results to rank models overall.

1Closed

Muse Spark 1.1

Meta

muse-spark-1-1

54.7%
Overall —Context 1M
2Closed

Claude Opus 4.7 (Adaptive)

Anthropic

claude-opus-4-7-max

45.9%
Overall 76Context 1M
3Closed

GPT-5.5

OpenAI

gpt-5-5

42.7%
Overall 80Context 1M
21 modelsAgenticCurrentDisplay onlyUpdated July 9, 2026

The published JobBench snapshot is tightly clustered at the top: Muse Spark 1.1 sits at 54.7%, while the third row is only 12.0 points behind. The broader top-10 spread is 27.0 points, so the benchmark still separates strong models even when the leaders cluster.

21 models have been evaluated on JobBench. The benchmark falls in the Agentic category. This category carries a 22% weight in BenchLM.ai's overall scoring system. JobBench is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About JobBench

Year

2026

Tasks

130 tasks across 35 occupations

Format

Agentic workplace deliverables

Difficulty

Professional multi-source workflows

JobBench evaluates agents on source-grounded workplace tasks across 35 occupations, using heterogeneous reference files and fact-anchored chained rubrics. BenchLM tracks the main-set score as a display-only professional-agent benchmark.

JobBench: Aligning Agent Work With Human Will

BenchLM freshness & provenance

Version

JobBench 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.

Benchmark score table (21 models)

1
Muse Spark 1.1muse-spark-1-1
54.7%
2
Claude Opus 4.7 (Adaptive)claude-opus-4-7-max
45.9%
3
GPT-5.5gpt-5-5
42.7%
4
GPT-5.4gpt-5-4
38.9%
5
Claude Sonnet 4.6claude-sonnet-4-6
36.9%
6
Claude Opus 4.6claude-opus-4-6
36.7%
7
GPT-5.2gpt-5-2
34.3%
8
GPT-5.3 Codexgpt-5-3-codex
33.7%
9
Claude Opus 4.5claude-opus-4-5
32.3%
10
Claude Sonnet 4.5claude-sonnet-4-5
27.7%
11
GPT-5.1-Codexgpt-5-1-codex
26.2%
12
GPT-5.2-Codexgpt-5-2-codex
26.0%
13
Claude 4.1 Opusclaude-4-1-opus
21.9%
14
Qwen3.5 Plusqwen3-5-plus
18.5%
15
Claude 4 Sonnetclaude-4-sonnet
18.4%
16
Claude Haiku 4.5claude-haiku-4-5
16.0%
17
MiniMax M2.5minimax-m2-5
14.8%
18
Gemini 3 Progemini-3-pro
11.4%
19
Gemini 3 Flashgemini-3-flash
11.4%
20
Kimi K2.5kimi-k2-5
8.7%
21
GPT-5 (high)gpt-5-high
8.5%

FAQ

What does JobBench measure?

An occupational agent benchmark for professional workflows that workers say they most want delegated to AI.

Which model scores highest on JobBench?

Muse Spark 1.1 by Meta currently leads with a score of 54.7% on JobBench.

How many models are evaluated on JobBench?

21 AI models have been evaluated on JobBench on BenchLM.

Compare Top Models on JobBench

Muse Spark 1.1 vs Claude Opus 4.7 (Adaptive)Claude Opus 4.7 (Adaptive) vs GPT-5.5GPT-5.5 vs GPT-5.4GPT-5.4 vs Claude Sonnet 4.6
Last updated: July 9, 2026 · BenchLM version JobBench 2026

The AI models change fast. We track them for you.

For engineers, researchers, and the plain curious — a weekly brief on new models, ranking shifts, and pricing changes.

Free. No spam. Unsubscribe anytime.