Gemini 4 Hydrogen vs Step 5 Preview

Change models
IndexAll categories, out of 100.
77.3 ±8.0 best
68.2 ±3.7
GapIndex points below the top.
Top
−9.1
CoverageIndex weight that has results.
65%
90% best
SpeedOutput tokens per second.
—
105/s
EloLMArena rating and rank.
—
1443
ReleasedMonth of release.
2026-11
2026-09
Input / 1MUS dollars per 1M input tokens.
$1.50
$1.00 best
Output / 1MUS dollars per 1M output tokens.
$7.50
$2.70 best
1K chats2K tokens in, 500 out, each.
$6.75
$3.35 best
1K documents20K tokens in, 1K out, each.
$37.50
$22.70 best
1K agent tasks200K tokens in, 20K out, each.
$450.00
$254.00 best
ContextMaximum tokens in one request.
1M
1M
AgenticMulti-step tasks with tools.
78.7 best
68.7
CodingCode writing and repair.
77.4 best
71.3
ReasoningLogic problems and puzzles.
—
67.2
MultimodalTasks with images and text.
80.6 best
62.1
KnowledgeFacts and expert knowledge.
74.3 best
63.4
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
—
—
MathMath problems.
—
61.5

Benchmarks

7 shared
Only Gemini 4 Hydrogen: 9Only Step 5 Preview: 55
Gemini 4 Hydrogen
Step 5 Preview
50.5% best19 / 165
36.4%65 / 165
87.8% best12 / 124
70.4%49 / 124
53.8% best6 / 91
40.4%36 / 91
52.3% best1 / 83
27.3%40 / 83
54.0% best10 / 62
31.8%24 / 62
11.3%19 / 36
12.9% best15 / 36
53.4% best3 / 20
27.6%20 / 20