Gemini 4 Hydrogen (xhigh): benchmark scores, price and ranking

Gemini 4 Hydrogen (xhigh) from Google DeepMind ranks #25 of 827 LLMs on the LLMs Tiger aggregate benchmark score (54.7), based on 2 benchmark results captured up to 2026-11-18.

Organisation
Google DeepMind
Type
Frontier model
Open weights
No
Released
2026-11-12 (source: epoch)
Cheapest API price
$1.50 in / $7.50 out per million tokens (via gemini, observed 2026-11-18T15:47Z)
Aggregate rank
#25 of 827 (score 54.7)

Across the 2 benchmarks where at least ten models are scored, Gemini 4 Hydrogen (xhigh) is in the top 10% on 0 and in the top half on 2.

Its cheapest listed input price, $1.50 per million tokens, is lower than that of 25% of the 443 models with a listed price.

Gemini 4 Hydrogen (xhigh) benchmark results

Gemini 4 Hydrogen (xhigh) benchmark scores, ranked against all models measured on each benchmark
BenchmarkScoreRankSourceCaptured
ALE-Bench1,826#11 of 93Epoch AI Benchmarking Hub2026-11-18
CursorBench57.8%#4 of 22Epoch AI Benchmarking Hub2026-11-18

About Gemini 4 Hydrogen (xhigh)

Gemini 4 Hydrogen (xhigh) is a frontier model from Google DeepMind. It was released on 2026-11-12, according to Epoch AI. The cheapest API listing we track charges $1.50 per million input tokens and $7.50 per million output tokens, through gemini. We hold 2 benchmark results for it across 1 area, and none is in the top tenth of its board.

Where Gemini 4 Hydrogen (xhigh) stands

On our aggregate score, Gemini 4 Hydrogen (xhigh) is #25 of 827 ranked models, in the top fifty. Its score is 54.7; the scale is centred on 50, so it sits 4.7 points above the typical model. That beats 97% of the ranked field. Among the 44 ranked Google DeepMind models it is #2. The next one up is Gemini 4 Carbon (high) at 62.4. The next one down is Gemini 3.8 Flash (high) at 53.4.

Gemini 4 Hydrogen (xhigh), benchmark by benchmark

ALE-Bench puts Gemini 4 Hydrogen (xhigh) at 1,826, #11 of 93. That beats 88% of the models tested, but 1,126 points short of GPT-6 Astra (max) (2,951). The results either side are Kimi K3.1 Pro (max) (1,933) and Qwen3.8 Max (0902) (xhigh) (1,679). It covers AtCoder heuristic-optimisation contests (performance rating). The board median is 800, so Gemini 4 Hydrogen (xhigh) is 1,026 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-11-18.

On CursorBench, Gemini 4 Hydrogen (xhigh) reaches 57.8%. It ranks #4 of 22, ahead of 82% of the field, and 2.6 points behind Gemini 4 Carbon (high) at 60.4%. The results either side are Claude Opus 5.5 (max) (57.8%) and Claude Sonnet 5.5 (max) (55.5%). It covers real-world coding-agent tasks. The board median is 46.5%, so Gemini 4 Hydrogen (xhigh) is 11.3 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-11-18.

Strengths and weaknesses

Software engineering: strong, ahead of 85% of the field on average across 2 benchmarks. Its best board here is ALE-Bench (#11 of 93) and its weakest is CursorBench (#4 of 22).

Price and value

The median input price across the 443 priced models is $0.482 per million tokens, so Gemini 4 Hydrogen (xhigh) costs more than the typical listing. Models that score within 1.5 points or better and cost less: DeepSeek V4.1 Flash 1017 (max) ($0.036, 55.5), Claude Haiku 5.5 (max) ($0.100, 54.3), Kimi K3 (max) ($0.192, 55.2). Similar scores cost more on GPT-5.3 Codex (xhigh) ($1.75) and GPT-5.6 Terra (max) ($2.00).

Release history

The Google DeepMind model released before it was Gemini 4 Carbon (high) on 2026-10-28, which scores 62.4.

Reading these numbers

The 2 scores on this page come from 1 source: Epoch AI Benchmarking Hub (2). No official leaderboard supplied any of them, so treat them as third-party measurements. The leaderboard also lists 4 other reasoning-effort settings of this model, which redirect here. This page shows the highest. With 2 results the aggregate rank can move a lot when one new score arrives.

Gemini 4 Hydrogen (xhigh) is also listed at 4 other reasoning-effort settings; this page shows the highest-effort row, which is the one the leaderboard keeps by default.

Models ranked nearby

More from Google DeepMind

Compare it on the live leaderboard → · How the aggregate score works