# Google Gemini 4 Hydrogen As of November 14, 2026, Google Gemini 4 Hydrogen ranks #12 of 118 models overall with 77.1 ±5.3 on the LLMBenchmark Index. Gemini 4 Hydrogen ranks #8 of 134 in web design, its strongest category. Google Gemini 4 Hydrogen costs $1.50 per million input tokens and $7.50 per million output tokens. Gemini 4 Hydrogen was released on November 12, 2026. Page: https://llmbenchmark.ai/models/gemini-4-hydrogen Data from November 14, 2026. Every figure below names the board it was read from and the day. - Maker: Google - Released: November 12, 2026 - Price per million input tokens: $1.50 - Price per million output tokens: $7.50 - Context window: 1.05M tokens - Overall rank: 12 of 118, by the LLMBenchmark Index - LLMBenchmark Index: 77.1 ±5.3 (90% margin, from 3 of 7 inputs) - Vibe Coding Index: 77.9 (from 4 of 6 coding benchmarks) Benchmark results (50 of 222): - LiveBench coding: 78.5% (source: LiveBench, read November 14, 2026) - LMArena WebDev: 1704 Elo (High) (source: LMArena WebDev, read November 14, 2026) - Vals Index: 68.5% (source: Vals Index, read November 14, 2026) - Design Arena: websites: 1320 Elo (source: Design Arena, read November 14, 2026) - Design Arena: game dev: 1369 Elo (source: Design Arena, read November 14, 2026) - Design Arena: 3D: 1415 Elo (source: Design Arena, read November 14, 2026) - Design Arena: all code categories: 1333 Elo (source: Design Arena, read November 14, 2026) - LiveBench reasoning: 88.6% (source: LiveBench, read November 14, 2026) - LiveBench data analysis: 83.2% (source: LiveBench, read November 14, 2026) - LiveBench research: 86.8% (source: LiveBench, read November 14, 2026) - LiveBench browser: 77.1% (source: LiveBench, read November 14, 2026) - LiveBench long agentic: 61.8% (source: LiveBench, read November 14, 2026) - LiveBench design: 77.3% (source: LiveBench, read November 14, 2026) - Terminal-Bench 4 (Vals AI): 54.0% (Mini-SWE-agent) (source: Vals AI boards, read November 14, 2026) - Vibe Code Bench (Vals AI): 87.8% (OpenHands) (source: Vals AI boards, read November 14, 2026) - Tax Agent Bench (Vals AI): 52.3% (source: Vals AI boards, read November 14, 2026) - Legal Research Bench (Vals AI): 53.8% (source: Vals AI boards, read November 14, 2026) - Harvey's Legal Agent Benchmark (Vals AI): 11.3% (source: Vals AI boards, read November 14, 2026) - Terminal-Bench 4.0 (Artificial Analysis run): 54.0% (High) (source: Artificial Analysis, read November 14, 2026) - FrontierCode 2.0: 39.0% (xHigh) (source: Artificial Analysis, read November 14, 2026) - HLE-Diamond (Artificial Analysis run): 58.8% (xHigh) (source: Artificial Analysis, read November 14, 2026) - AA-Omniscience: 46.3 (xHigh) (source: Artificial Analysis, read November 14, 2026) - CritPt v2, hard subset (Artificial Analysis run): 66.0% (xHigh) (source: Artificial Analysis, read November 14, 2026) - LMArena Text: 1502 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: coding: 1516 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: hard prompts: 1520 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: math: 1507 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: instruction following: 1519 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: creative writing: 1509 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: English: 1499 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: non-English: 1495 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: Chinese: 1631 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: French: 1512 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: Spanish: 1494 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: Russian: 1492 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: multi-turn: 1485 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: longer queries: 1524 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: expert: 1558 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: hard prompts in English: 1514 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: software and IT: 1513 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: mathematical work: 1534 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: science: 1519 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: medicine and healthcare: 1488 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: legal and government: 1532 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: business and finance: 1486 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: writing and language: 1500 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena Text: entertainment and media: 1503 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena WebDev: React: 1724 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - LMArena WebDev: HTML: 1743 Elo (High) (source: LMArena (Hugging Face dataset), read November 14, 2026) - CursorBench 4.0: 57.8% (xhigh) (source: CursorBench (Cursor), read November 14, 2026)