APEX-Agents
We show this table for reference; we do not rank on it.
A professional-services agent benchmark covering long-horizon knowledge-work tasks.
Benchmark score on APEX-Agents — November 13, 2026
We compile the APEX-Agents rows from provider self-reports and secondary reports. Hy4 leads the table at 75.2%, followed by Kimi K3.1 Pro (69.4%) and ERNIE 5.5 Flash Preview (64.2%). We do not use these results to rank models overall.
Hy4
Tencent
Reported by TencentKimi K3.1 Pro
Moonshot AI
Reported by Moonshot AIERNIE 5.5 Flash Preview
Baidu
Reported by Baidu13 modelsAgenticCurrentDisplay onlyUpdated November 13, 2026
Benchmark score table (13 models)
None of the 13 results on this page has a documented run by the benchmark or an independent evaluator: 12 reported by the labs that built the models and 1 reported in another publisher’s document.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | Hy4TencentReported by Tencent | 75.2% | Reported by Tencent | Not reported | Pending |
| 2 | Kimi K3.1 ProMoonshot AIReported by Moonshot AI | 69.4% | Reported by Moonshot AI | Not reported | Open |
| 3 | ERNIE 5.5 Flash PreviewBaiduReported by Baidu | 64.2% | Reported by Baidu | Not reported | Pending |
| 4 | Grok 4.6xAIReported by xAI | 57.5% | Reported by xAI | Not reported | Closed |
| 5 | Apodex 1.1ApodexReported by Apodex | 38.5% | Reported by Apodex | Not reported | Closed |
| 6 | Step 5 PreviewStepFunReported by StepFun | 37.8% | Reported by StepFun | Not reported | Open |
| 7 | Kimi K3Moonshot AIReported by Moonshot AI | 37.6% | Reported by Moonshot AI | Not reported | Open |
| 8 | Hy4 previewTencentReported by Tencent | 37.1% | Reported by Tencent | Not reported | Open |
| 9 | dots3-note PreviewDots StudioReported by Dots Studio | 30.8% | Reported by Dots Studio | Not reported | Open |
| 10 | Ling 3.0 Flash FinInclusionAIReported by a third party | 29.2% | Reported by a third party | Not reported | Closed |
| 11 | Apodex 1.1 MiniApodexReported by Apodex | 27.7% | Reported by Apodex | Not reported | Open |
| 12 | Solar Pro 4UpstageReported by Upstage | 18.7% | Reported by Upstage | Not reported | Closed |
| 13 | Solar Open 2UpstageReported by Upstage | 16.6% | Reported by Upstage | Not reported | Open |
- Lab-reported
- Benchmark-run
- Independent
Results reported by model developersUnranked
Labs ran these themselves, with a metric, split or setting the table above does not use. They take no rank here.
| Model | Reported value | What was measured | Against the table above | Who ran it |
|---|---|---|---|---|
| Apodex 1.1Apodex | 34.4Score | ReAct single agentThe stored row uses Agent Team | Not comparable to the scores above | Reported by Apodex |
| Apodex 1.1 MiniApodex | 24.2Score | ReAct single agentThe stored row uses Agent Team | Not comparable to the scores above | Reported by Apodex |
A lab’s own result is shown for reference. It never takes a rank on this board.
How we label sourcesAmong the reported APEX-Agents rows, Hy4 is first at 75.2%. The third row is 11.0 points behind. The broader top-10 range is 46.0 points, so the table still separates the published systems.
13 models have been evaluated on APEX-Agents. The benchmark falls in the Agentic category. APEX-Agents is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About APEX-Agents
Year
2026
Tasks
Professional-services agent tasks
Format
Agent task-completion score
Difficulty
Long-horizon professional work
BenchLM stores Moonshot's provider-run APEX-Agents value separately from the Artificial Analysis APEX-Agents-AA implementation.
Freshness and provenance
Version
APEX-Agents 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
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 APEX-Agents measure?
A professional-services agent benchmark covering long-horizon knowledge-work tasks.
Which model scores highest on APEX-Agents?
Hy4 by Tencent currently leads with a score of 75.2% on APEX-Agents.
How many models are evaluated on APEX-Agents?
13 AI models have published results on APEX-Agents in the BenchLM catalog.
Compare top models on APEX-Agents
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.