Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Coding benchmark
Every model the catalog carries a published OpenHarmony Bench v1.0 value for, ranked by that value.
An app-level coding benchmark that asks DevEco Code configurations to implement observable behavior in buildable OpenHarmony ArkTS applications. Replaced by OpenHarmony Bench v1.2 in November 2026. 14 models in the ModelScale catalog have a published value.
Each point is a model with a published OpenHarmony Bench v1.0 value and complete token prices, on the 75% input / 25% output blend. Frontier models are spaced evenly along the cost axis.
14 models with a published OpenHarmony Bench v1.0 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 | Task completion through DevEco Code |
|---|---|---|---|
| 1 | Z.AI | 60.8 | |
| 1 | Alibaba | 60.8 | |
| 3 | DeepSeek | 60.3 |
What the benchmark measures and how its value is published, from the source’s own record.
OpenHarmony Bench v1.0 is a coding benchmark published by OpenHarmony Bench: Evaluating LLMs and Coding Agents on OpenHarmony App Development. An app-level coding benchmark that asks DevEco Code configurations to implement observable behavior in buildable OpenHarmony ArkTS applications. Replaced by OpenHarmony Bench v1.2 in November 2026. Its evaluation set is described by the source as 153 app-development and bug-fix tasks. The source rates its difficulty as end-to-end OpenHarmony application development.
Each model's result is published by OpenHarmony Bench: Evaluating LLMs and Coding Agents on OpenHarmony App Development as task completion through DevEco Code. ModelScale prints that value to 1 decimal place, exactly as published, and never converts or rescales it. This page ranks the 14 models that carry a published value, highest value first; a model the source never scored is left out rather than ranked at zero. The source does not state whether a higher value is the better result, so neither does this page.
| 4 | DeepSeek | 59.0 |
| 4 | DeepSeek | 59.0 |
| 6 | Z.AI | 58.4 |
| 7 | Z.AI | 57.3 |
| 7 | Moonshot AI | 57.3 |
| 9 | Alibaba | 56.0 |
| 10 | DeepSeek | 53.8 |
| 11 | Alibaba | 53.4 |
| 12 | Z.AI | 52.3 |
| 13 | Moonshot AI | 52.1 |
| 14 | MiniMax | 48.4 |
Evidence key: ObservedLast good
Rows are ordered by the value OpenHarmony Bench: Evaluating LLMs and Coding Agents on OpenHarmony App Development published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.