OpenAI’s GPT-5.5 AI Model Secures the Top Position on Android Bench Leaderboard
OpenAI has recently released its latest AI model, GPT-5.5, and as per the updated Android Bench score, this new model has secured the top position on the AI benchmark leaderboard.
Check out which LLM model has secured which position below.
Before getting into the scoreboard, Android Bench is an official framework and leaderboard developed by Google, designed to evaluate how well large language models (LLMs) handle real-world Android development tasks. It tests AI models on codebases, assessing their ability to generate accurate patches, resolve API issues, and handle tasks such as Jetpack Compose migration, thereby helping developers choose suitable AI tools.
As per the details shared, OpenAI’s latest GPT-5.5 has secured the top position with a confidence interval (CI) of 66.8%–80.5%, successfully resolving an average of 74.0% of the tasks. This is followed by a tie for the next highest score: OpenAI’s GPT-5.4 and Google’s own Gemini 3.1 Pro Preview. Both of these models have successfully resolved an average of 72.4% of the tasks, with a CI of 65.4%–79.3% and 65.1%–78.8%, respectively. Standing at the fourth position is Claude Opus 4.7, which has resolved an average of 68.7% of the tasks with a CI of 61.2%–76.0%.
Below is the list of AI models’ scores on the Android Bench as of May 5th 2026
| Model | Score | Range | Date |
|---|---|---|---|
| GPT 5.5 | 74.0 | 66.8 — 80.5 | 2026-04-27 |
| GPT 5.4 | 72.4 | 65.4 — 79.3 | 2026-03-16 |
| Gemini 3.1 Pro Preview | 72.4 | 65.1 — 78.8 | 2026-02-27 |
| Claude Opus 4.7 | 68.7 | 61.2 — 76.0 | 2026-04-27 |
| GPT 5.3 Codex | 67.7 | 59.9 — 75.1 | 2026-03-18 |
| Claude Opus 4.6 | 66.6 | 59.5 — 73.9 | 2026-02-26 |
| GPT 5.2 Codex | 62.5 | 54.6 — 70.1 | 2026-02-26 |
| Claude Opus 4.5 | 61.9 | 53.0 — 70.1 | 2026-02-26 |
| Gemini 3 Pro Preview | 60.4 | 52.3 — 68.2 | 2026-02-27 |
| Claude Sonnet 4.6 | 58.4 | 50.4 — 66.5 | 2026-02-27 |
| Claude Sonnet 4.5 | 53.8 | 45.5 — 62.2 | 2026-02-26 |
| Gemini 3 Flash Preview | 42.0 | 36.5 — 47.6 | 2026-02-26 |
| Gemini 2.5 Flash | 16.7 | 11.5 — 22.1 | 2026-02-26 |
