Google DeepMindproprietary
Released: 2026-11-12Artificial AnalysisGemini 4 Hydrogen
Gemini 4 Hydrogen is the high-efficiency workhorse model of the Gemini 4 family from Google, built for high-throughput agentic workflows, coding and multi-model routing at lower cost and latency than...
Intelligence Index
53.9/ 100
Artificial Analysis Intelligence Index (independent composite)Throughput Speed
174tok/sec
Output streaming throughputTime to First Answer
47s
Time to first answer token, including any thinkingCost to Run the Index
3.16USD
Artificial Analysis cost of the full Index runMeasured results · Artificial Analysis
Independent evaluations of Gemini 4 Hydrogen (Xhigh configuration). Values are reported as published; nothing here is estimated or derived. Snapshot 2026-11-15.
- Intelligence Index
- 53.9
- Cost to run the Index
- $3.16
- Time per task
- 14.8 min
- Omniscience index
- 46.3
- Reasoning model
- Yes
- Cached-input price
- $0.075 / 1M
Reasoning & knowledge
- AA-Omniscience Accuracy50.5%
- CritPt v2 (Hard)66.0%
- HLE-Diamond58.8%
Coding & terminal agents
Agentic work
- AA-Briefcase v1.1 (Elo)1568 Elo
- AutomationBench-AA76.1%
Long context & documents
- GDP.pdf27.6%
- MLCR-AA (medical long context)21.1%
Health & safety
Reasoning-effort variants evaluated
| Variant | Intelligence Index | Cost to run Index | Output speed (tok/s) |
|---|---|---|---|
| Xhigh | 53.9 | $3.16 | 174.0 |
| High | 50.8 | $1.87 | 169.3 |
| Medium | 41.2 | $0.80 | 172.8 |
| Low | 26.9 | $0.29 | 166.9 |
Headline Evaluation Breakdown
Six independently measured evaluations (Artificial Analysis)Reasoning · CritPt v2 (Hard)66.0%
Coding & agents · Terminal-Bench 4.053.5%
Sourced Benchmark Evaluations (10)
Verified performance across authoritative benchmarks with provenance tracking.| Benchmark | Status | Score | Trust | Date | Source Type | Provenance |
|---|---|---|---|---|---|---|
| Terminal-Bench 4.0 | Active | 53.5% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench-Science 0.1 | Active | 49.5% | 93 | 2026-11-15 | independent | Source ↗ |
| CritPt v2 (Hard) | Active | 66% | 93 | 2026-11-15 | independent | Source ↗ |
| HLE-Diamond | Active | 58.8% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-Omniscience (Accuracy) | Active | 50.5% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-Omniscience (Non-Hallucination) | Active | 91.5% | 93 | 2026-11-15 | independent | Source ↗ |
| FrontierCode-AA 2.0 | Active | 39% | 93 | 2026-11-15 | independent | Source ↗ |
| GDP.pdf | Active | 27.6% | 93 | 2026-11-15 | independent | Source ↗ |
| MLCR-AA (Medical Long Context Reasoning) | Active | 21.1% | 93 | 2026-11-15 | independent | Source ↗ |
| AutomationBench-AA | Active | 76.1% | 93 | 2026-11-15 | independent | Source ↗ |
OpenAI Compatible API ExecutionAPI Ready
from openai import OpenAI
client = OpenAI()
response = client.chat.completions.create(
model="google-gemini-4-hydrogen",
messages=[{"role": "user", "content": "Evaluate multi-step logic problem"}],
temperature=0.2
)
print(response.choices[0].message.content)Evaluation & Sourcing Notes
Scores listed for Gemini 4 Hydrogen represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.
All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.