Coding · Benchmark profile
An open benchmark for AI coding agents on real-world React Native implementation tasks, emphasizing working app behavior, recommended architecture choices, and strict constraint adherence.
Data verified 27 Jul 2026 · Methodology 1.8.0
Benchmark score on React Native Evals
Composer 2 leads at 96.1 elo, followed by Composer 2 Fast (94.9 elo) and GPT-5.4 (85.3 elo).
10 ranked models · higher is better · labels show rank and score
| Model | Rank | Provider | Score |
|---|---|---|---|
| Composer 2 | #1 | Cursor | 96.1 elo |
| Composer 2 Fast | #2 | Cursor | 94.9 elo |
| GPT-5.4 | #3 | OpenAI | 85.3 elo |
| GPT-5.5 | #4 | OpenAI | 84.7 elo |
| Claude Opus 4.6 | #5 | Anthropic | 84.1 elo |
| Claude Opus 4.7 | #6 | Anthropic | 82.8 elo |
| Claude Sonnet 4.6 | #7 | Anthropic | 80.6 elo |
| Gemini 3.1 Pro | #8 | 78.9 elo | |
| Kimi K2.5 | #9 | Moonshot AI | 77.2 elo |
| Gemma 4 31B | #10 | 75.2 elo |
One best score per model · higher is better
| Rank | Model | Provider | License | Evidence use | Score |
|---|---|---|---|---|---|
| #1 | Composer 2 composer-2 | Cursor | closed | Estimated reference | 96.1 elo |
| #2 | Composer 2 Fast composer-2-fast | Cursor | closed | Estimated reference | 94.9 elo |
| #3 | GPT-5.4 gpt-5.4 | OpenAI | closed | Reference only | 85.3 elo |
| #4 | GPT-5.5 gpt-5.5 | OpenAI | closed | Estimated reference | 84.7 elo |
| #5 | Claude Opus 4.6 claude-opus-4-6 | Anthropic | closed | Reference only | 84.1 elo |
| #6 | Claude Opus 4.7 claude-opus-4-7 | Anthropic | closed | Reference only | 82.8 elo |
| #7 | Claude Sonnet 4.6 claude-sonnet-4-6 | Anthropic | closed | Reference only | 80.6 elo |
| #8 | Gemini 3.1 Pro gemini-3-1-pro | closed | Estimated reference | 78.9 elo | |
| #9 | Kimi K2.5 kimi-k2.5 | Moonshot AI | open | Reference only | 77.2 elo |
| #10 | Gemma 4 31B gemma-4-31b | open | Reference only | 75.2 elo | |
| #11 | GLM-5 glm-5 | Z.AI | open | Reference only | 74.8 elo |
| #12 | Grok 4 grok-4 | xAI | closed | Reference only | 72.6 elo |
| #13 | GPT-OSS 120B gpt-oss-120b | OpenAI | open | Reference only | 71.6 elo |
| #14 | DeepSeek V3.2 deepseek-v3.2 | DeepSeek | open | Reference only | 71.5 elo |
| #15 | MiniMax M2.7 minimax-m2-7 | MiniMax | open | Reference only | 71.4 elo |
| #16 | GPT-OSS 20B gpt-oss-20b | OpenAI | open | Reference only | 71 elo |
The top of this snapshot is led by Composer 2 at 96.1 elo; third place is 10.8 points behind. The top-10 spread is 20.9 points.
About React Native Evals
An open benchmark for AI coding agents on real-world React Native implementation tasks, emphasizing working app behavior, recommended architecture choices, and strict constraint adherence. Results stay tied to the exact model variant and evaluation system. Multiple systems for the same model use the best published score on this page; overall Lumina scoring uses the median of ranking-eligible rows.
Open benchmark source ↗FAQ
An open benchmark for AI coding agents on real-world React Native implementation tasks, emphasizing working app behavior, recommended architecture choices, and strict constraint adherence.
Composer 2 by Cursor currently leads with 96.1 elo.
16 models in the LuminaBench cohort have a qualifying score on this benchmark.
Related
