AI Coding Daily (OpenCode) - React-TS Code Quality: leaderboard
React-TS project code quality on AI Coding Daily's LLM Coding Leaderboard for runs in the OpenCode harness: a 100-point rubric score from a fixed LLM judge (gpt-5.6-sol at medium effort in Codex CLI) divided by five, maximum 20; higher is better. The judge is an OpenAI model, so self-preference is possible.
Metric: React-TS Code Quality (max 20) points, LLM-judged rubric score / 5. Source: aicodingdaily.com. Saturation forecast: Estimated already saturated. 22 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | mistral-large-4-oss (High) | 18.5 | #60 (mistral-large-4-oss) |
| 2 | Muse Spark 1.3 (Max) | 18.33 | #31 (Muse Spark 1.3) |
| 3 | Kimi K3.1 Pro (Max) | 18.33 | #17 (Kimi K3.1 Pro) |
| 4 | DeepSeek V4.1 Flash (1017) (Max) | 18.17 | #27 (DeepSeek V4.1 Flash (1017)) |
| 5 | GLM-5.3 (High) | 17.67 | #44 (GLM-5.3) |
| 6 | Qwen 3.8 Flash (Max) | 17.67 | #62 (Qwen 3.8 Flash) |
| 7 | Kimi K3.1 Pro (High) | 17.33 | #17 (Kimi K3.1 Pro) |
| 8 | GLM-5.3 Flash (Max) | 17.17 | #77 (GLM-5.3 Flash) |
| 9 | MiniMax-M3.1-Flash (Max) | 17 | #52 (MiniMax-M3.1-Flash) |
| 10 | DeepSeek V4.1 Flash (1017) (High) | 16.83 | #27 (DeepSeek V4.1 Flash (1017)) |
| 11 | DeepSeek V4 Pro (0813) (Max) | 16.5 | #58 (DeepSeek V4 Pro (0813)) |
| 12 | DeepSeek V4 Pro (0813) (High) | 16.33 | #58 (DeepSeek V4 Pro (0813)) |
| 13 | Mercury 3 (High) | 15.83 | #78 (Mercury 3) |
| 14 | Qwen 3.8 27B (xHigh) | 15.67 | #123 (Qwen 3.8 27B) |
| 15 | Gemini 3.1 Pro (Preview) | 15.33 | #64 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=ai-coding-daily-opencode-react-ts-code-quality · How It Works · Data refreshed daily, snapshot 2026-11-05.