{"slug": "gemini-4-neon-is-the-new-1-on-the-models-engineering-index", "title": "Gemini 4 Neon is the new #1 on the Models Engineering Index", "summary": "Google's Gemini 4 Neon ranks first on the Models Engineering Index (Mei), which scores language models on practical engineering tasks, with an index score of 69.5. Anthropic's Fable 5.2 places second at 65.8 and Anthropic's Opus 5.5 third at 63.1, while OpenAI's 6.2 Astra is fourth at 60.5. The ranking covers 19 models, with Meta's Wonder Full eighth at 49.3, Moonshot's K3.1 Pro ninth at 49.0 and xAI's Grok 4.7 fourteenth at 42.7.", "body_md": "# Mei—or, the Models Engineering Index—calculates language models' practical performance on engineering tasks. Mei calculates language models' engineering performance.\n\nPerformance versus .\n\n## Rankings\n\n|  |  | Index bar |  |  |  | \n|---|---|---|---|---|---|\n| 1 | 4 Neon Google |  | 69.5 | 1.9× | 1× | \n| 2 | Fable 5.2 Anthropic |  | 65.8 | 6.5× | 2.1× | \n| 3 | Opus 5.5 Anthropic |  | 63.1 | 3.5× | 1.7× | \n| 4 | 6.2 Astra OpenAI |  | 60.5 | 1.4× | 0.29× | \n| 5 | 4 Hydrogen Google |  | 57.2 | 1.2× | 1.1× | \n| 6 | Sonnet 5.5 Anthropic |  | 57.2 | 2.5× | 2.5× | \n| 7 | 6.2 Sol OpenAI |  | 56.9 | 0.4× | 0.35× | \n| 8 | Wonder Full Meta |  | 49.3 | 1.1× | 0.49× | \n| 9 | K3.1 Pro Moonshot |  | 49.0 | 1.2× | 0.74× | \n| 10 | 4-Max Alibaba |  | 49.0 | 7.3× | 1.2× | \n| 11 | 6.1 Terra OpenAI |  | 47.2 | 0.83× | 1.7× | \n| 12 | Wonder Meta |  | 46.4 | 1.3× | 0.63× | \n| 13 | V4.1 Flash DeepSeek |  | 44.1 | 0.13× | 1.4× | \n| 14 | Grok 4.7 xAI |  | 42.7 | 3.9× | 0.9× | \n| 15 | Haiku 5.5 Anthropic |  | 42.5 | 0.4× | 1.6× | \n| 16 | GLM 5.3 Z.ai |  | 42.0 | 2.2× | 1.5× | \n| 17 | Hy4 Tencent |  | 37.3 | 0.33× | 1.1× | \n| 18 | V2.6-Pro Xiaomi |  | 35.4 | 0.13× | 0.93× | \n| 19 | 6 Luna OpenAI |  | 34.8 | 0.065× | 0.62× | \n\nComparing and", "url": "https://wpnews.pro/news/gemini-4-neon-is-the-new-1-on-the-models-engineering-index", "canonical_source": "https://git-aarvin.github.io/mei/", "published_at": "2026-11-19 23:35:35+00:00", "updated_at": "2026-11-19 23:44:32.257556+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-research"], "entities": ["Google", "Gemini 4 Neon", "Models Engineering Index", "Anthropic", "Fable 5.2", "Opus 5.5", "OpenAI", "6.2 Astra"], "also_reported_by": [], "takeaway": "Gemini 4 Neon tops the Models Engineering Index at 69.5, ahead of Anthropic's Fable 5.2 at 65.8 and Opus 5.5 at 63.1, with OpenAI's 6.2 Astra fourth at 60.5.", "recommendation": {"advice": "Benchmark candidate models on engineering tasks with a task-specific index like Mei rather than general leaderboards; Gemini 4 Neon scored 69.5 versus 65.8 for Fable 5.2.", "playbook": "evals-testing", "playbook_url": "https://wpnews.pro/playbooks/evals-testing"}, "alternates": {"html": "https://wpnews.pro/news/gemini-4-neon-is-the-new-1-on-the-models-engineering-index", "markdown": "https://wpnews.pro/news/gemini-4-neon-is-the-new-1-on-the-models-engineering-index.md", "text": "https://wpnews.pro/news/gemini-4-neon-is-the-new-1-on-the-models-engineering-index.txt", "jsonld": "https://wpnews.pro/news/gemini-4-neon-is-the-new-1-on-the-models-engineering-index.jsonld"}}