Hy4 flash vs Qwen3.8 Max

Change models
Hy4 flashTencent
IndexAll categories, out of 100.
65.1 ±9.2
65.1 ±2.5
GapIndex points below the top.
Top
Top
CoverageIndex weight that has results.
65%
95% best
SpeedOutput tokens per second.
72/s best
41/s
EloLMArena rating and rank.
—
1485
ReleasedMonth of release.
2026-11
2026-08
Input / 1MUS dollars per 1M input tokens.
$0.044
—
Output / 1MUS dollars per 1M output tokens.
$0.176
—
1K chats2K tokens in, 500 out, each.
$0.175
—
1K documents20K tokens in, 1K out, each.
$1.05
—
1K agent tasks200K tokens in, 20K out, each.
$12.26
—
ContextMaximum tokens in one request.
1.05M best
1M
AgenticMulti-step tasks with tools.
71.1 best
63.3
CodingCode writing and repair.
50.8
63.8 best
ReasoningLogic problems and puzzles.
—
66.4
MultimodalTasks with images and text.
53.6
66.8 best
KnowledgeFacts and expert knowledge.
79.1 best
62.3
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
—
71.1
MathMath problems.
—
67.2

Benchmarks

6 shared
Only Hy4 flash: 3Only Qwen3.8 Max: 108
Hy4 flash
Qwen3.8 Max
55.2% best15 / 79
43.6%31 / 79
DeepSWEAgentic
64.6% best36 / 61
56.6%48 / 61
75.8%34 / 54
82.3% best6 / 54
55.2%20 / 37
56.2% best15 / 37
52.2% best10 / 33
27.3%31 / 33
45.8%
49.0%