Gemini 4 Hydrogen
Gemini 4 Hydrogen ranks first on Tax Agent Bench.
23
Benchmarks with a score
1
Ranked first, out of five or more models
4
In the top three
Gemini 4 Hydrogen
Closed weights
No Epoch Capabilities Index score yet
- AutomationBench 51.1% #4 /42
- FrontierCode 2.0 38.9% #7 /32
- FrontierSWE 53.1% #8 /25
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Vibe Code Bench | 87.8% | #13 of 120 | Official leaderboard | |
| FrontierCode 2.0 | 38.9% | #7 of 32 | Official leaderboard | |
| CursorBench 4.0 | 54.3% | #5 of 21 | Official leaderboard | |
| SWE-Marathon v1.1 | 49.4% | #6 of 20 | Official leaderboard | |
| FrontierSWE | 53.1% | #8 of 25 | Official leaderboard |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| AutomationBench | 51.1% | #4 of 42 | Official leaderboard | |
| Terminal-Bench 4.1 | 56.4% | #8 of 43 | Official leaderboard |
Professional work
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Tax Agent Bench | 52.3% | #1 of 84 | Official leaderboard | |
| APEX-1 | 76.9% | #3 of 71 | Official leaderboard | |
| Vals Finance Index | 66.1% | #2 of 40 | Official leaderboard | |
| Legal Research Bench | 53.9% | #6 of 93 | Official leaderboard | |
| APEX-Agents 1.1 | 84% | #5 of 65 | Official leaderboard | |
| APEX-Accounting | 67.6% | #4 of 49 | Official leaderboard | |
| Excel Modeling Benchmark v1.1 | 75.8% | #6 of 42 | Official leaderboard | |
| Vals Index v2.2 | 68.5% | #6 of 37 | Official leaderboard | |
| Vals Legal Index | 73.6% | #9 of 38 | Official leaderboard | |
| Vals Finance Agent v2.1 | 65.1% | #12 of 40 | Official leaderboard | |
| Harvey LAB v1.1 (HLAB-Vals-1.1) | 11.3% | #21 of 38 | Official leaderboard |
Vision, audio and robotics
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MineBench | 1,993 | #6 of 75 | Official leaderboard |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| TxBench-Oligo-Discovery | 54.3% | #4 of 30 | Official leaderboard | |
| CWE-bench 2 | 68.1% | #2 of 14 | Official leaderboard | |
| TxBench-Antibody-Discovery | 57.7% | #5 of 30 | Official leaderboard | |
| TxBench-Preclinical-Pharmacology | 61.3% | #9 of 30 | Official leaderboard |