# APEX-Agents leaderboard

Agentic benchmark

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelscale.dev/leaderboards/benchmarks/agentic-apex-agents

Every model the catalog carries a published APEX-Agents value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Agentic |
| Measure | Agent task-completion score |
| Tasks | Professional-services agent tasks |
| Difficulty | Long-horizon professional work |
| Published by | [Kimi K3: Open Frontier Intelligence](https://www.kimi.com/blog/kimi-k3) |

A professional-services agent benchmark covering long-horizon knowledge-work tasks.

## APEX-Agents ranking

14 models with a published APEX-Agents value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

| Rank | Model | Provider | Agent task-completion score | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [Hy4](https://modelscale.dev/models/hy4) | Tencent | 75.2 | Observed 2026-11-21 · source benchlm:models |
| 2 | [Kimi K3.1 Pro](https://modelscale.dev/models/kimi-3-1-pro) | Moonshot AI | 69.4 | Observed 2026-11-21 · source benchlm:models |
| 3 | [ERNIE 5.5 Flash Preview](https://modelscale.dev/models/ernie-5-5-flash-preview) | Baidu | 64.2 | Observed 2026-11-21 · source benchlm:models |
| 4 | [Hy4 flash](https://modelscale.dev/models/hy4-flash) | Tencent | 63.1 | Observed 2026-11-21 · source benchlm:models |
| 5 | [Grok 4.6](https://modelscale.dev/models/grok-4-6) | xAI | 57.5 | Observed 2026-11-21 · source benchlm:models |
| 6 | [Apodex 1.1](https://modelscale.dev/models/apodex-1-1) | Apodex | 38.5 | Observed 2026-11-21 · source benchlm:models |
| 7 | [Step 5 Preview](https://modelscale.dev/models/step-5-preview) | StepFun | 37.8 | Observed 2026-11-21 · source benchlm:models |
| 8 | [Kimi K3](https://modelscale.dev/models/kimi-3) | Moonshot AI | 37.6 | Observed 2026-11-21 · source benchlm:models |
| 9 | [Hy4 preview](https://modelscale.dev/models/hy4-preview) | Tencent | 37.1 | Observed 2026-11-21 · source benchlm:models |
| 10 | [dots3-note Preview](https://modelscale.dev/models/dots3-note-preview) | Dots Studio | 30.8 | Observed 2026-11-21 · source benchlm:models |
| 11 | [Ling 3.0 Flash Fin](https://modelscale.dev/models/ling-3-0-flash-fin) | InclusionAI | 29.2 | Observed 2026-11-21 · source benchlm:models |
| 12 | [Apodex 1.1 Mini](https://modelscale.dev/models/apodex-1-1-mini) | Apodex | 27.7 | Observed 2026-11-21 · source benchlm:models |
| 13 | [Solar Pro 4](https://modelscale.dev/models/solar-pro-4) | Upstage | 18.7 | Observed 2026-11-21 · source benchlm:models |
| 14 | [Solar Open 2](https://modelscale.dev/models/solar-open-2) | Upstage | 16.6 | Observed 2026-11-21 · source benchlm:models |

Rows are ordered by the value the source published, highest first. The source does not state whether a higher value is the better result, so this document does not either: for a benchmark that measures a rate of failure, read the table from the bottom.
