Inference
Callstack

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.

ModelsModels with a result.16
Top resultBest result on this test.96.1%Composer 2
Top-3 spreadPoints from first to third.10.8 pts
YearYear of release.2026

Result and price

020406080100$0.1$0.3$1$3$10$30$100RESULTOUTPUT PRICE PER 1M TOKENS · LOG SCALE

Results

16 results
#ModelResultThe score from the source.
01Composer 296.1%
02Composer 2 Fast94.9%
03GPT-5.485.3%
04GPT-5.584.7%
05Claude Opus 4.684.1%
06Claude Opus 4.782.8%
07Claude Sonnet 4.680.6%
08Gemini 3.1 Pro78.9%
09Kimi K2.577.2%
10Gemma 4 31B75.2%
11GLM-574.8%
12Grok 472.6%
13GPT-OSS 120B71.6%
14DeepSeek V3.271.5%
15MiniMax M2.771.4%
16GPT-OSS 20B71.0%