Gemini 3.8 Flash vs Hy4 preview

Change models
Hy4 previewTencent
IndexAll categories, out of 100.
66.5 ±2.6 best
66.2 ±4.0
GapIndex points below the top.
Top
−0.3 · no clear gap
CoverageIndex weight that has results.
95% best
90%
SpeedOutput tokens per second.
91/s best
60/s
EloLMArena rating and rank.
1500
—
ReleasedMonth of release.
2026-09
2026-08
Input / 1MUS dollars per 1M input tokens.
$0.825 best
$0.834
Output / 1MUS dollars per 1M output tokens.
$4.13
$2.50 best
Batch output / 1MBatch price per 1M output tokens.
$1.88
—
1K chats2K tokens in, 500 out, each.
$3.71
$2.92 best
1K documents20K tokens in, 1K out, each.
$20.63
$19.18 best
1K agent tasks200K tokens in, 20K out, each.
$247.50
$216.82 best
ContextMaximum tokens in one request.
1.05M
1.05M
AgenticMulti-step tasks with tools.
63.0
63.1 best
CodingCode writing and repair.
65.0 best
64.9
ReasoningLogic problems and puzzles.
68.1 best
67.9
MultimodalTasks with images and text.
66.2
74.2 best
KnowledgeFacts and expert knowledge.
69.8 best
67.0
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
88.8
—
MathMath problems.
63.3
75.7 best

Benchmarks

29 shared
Only Gemini 3.8 Flash: 72Only Hy4 preview: 25
Gemini 3.8 Flash
Hy4 preview
18.3% best33 / 207
16.9%42 / 207
87.0% best7 / 149
83.8%47 / 149
78.7% best39 / 130
77.5%43 / 130
MedScribeHealthcare
84.5% best36 / 111
83.6%47 / 111
MedCodeHealthcare
48.1% best31 / 108
43.2%48 / 108
38.9%43 / 98
45.2% best25 / 98
32.1%26 / 89
32.6% best24 / 89
89.4% best5 / 89
85.4%18 / 89
61.4% best2 / 79
55.1%23 / 79
36.5%35 / 77
47.4% best14 / 77
EMBFinance
72.2% best8 / 75
58.1%38 / 75
81.3% best8 / 75
55.1%49 / 75
19.7% best39 / 68
8.1%54 / 68
1.0% best16 / 66
0.0%35 / 66
DeepSWEAgentic
73.8% best13 / 64
64.3%39 / 64
−2.837 / 59
−1.1 best31 / 59
1.033 / 59
3.2 best23 / 59
48.0%35 / 55
75.0% best19 / 55
18.2%35 / 52
19.1% best34 / 52
−0.735 / 50
6.5 best12 / 50