Inference
EN

Hy4 vs MiniMax M3.1 Flash

Change models
Hy4Tencent
MiniMax M3.1 FlashMiniMax
IndexAll categories, out of 100.
72.0 ±7.8 best
69.3 ±5.4
GapIndex points below the top.
Top
−2.7 · no clear gap
CoverageIndex weight that has results.
65%
80% best
SpeedOutput tokens per second.
52/s
100/s best
EloLMArena rating and rank.
—
—
ReleasedMonth of release.
2026-11
2026-10
Input / 1MUS dollars per 1M input tokens.
$0.417
$0.15 best
Output / 1MUS dollars per 1M output tokens.
$1.25
$0.60 best
1K chats2K tokens in, 500 out, each.
$1.46
$0.60 best
1K documents20K tokens in, 1K out, each.
$9.59
$3.60 best
1K agent tasks200K tokens in, 20K out, each.
$108.41
$42.00 best
ContextMaximum tokens in one request.
1.05M
1.05M
AgenticMulti-step tasks with tools.
74.5 best
71.8
CodingCode writing and repair.
68.4
69.4 best
ReasoningLogic problems and puzzles.
—
67.9
MultimodalTasks with images and text.
57.4
58.1 best
KnowledgeFacts and expert knowledge.
83.7 best
68.9
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
—
—
MathMath problems.
—
—

Benchmarks

10 shared
Only Hy4: 11Only MiniMax M3.1 Flash: 26
Hy4
MiniMax M3.1 Flash
Vibe Code Bench v1.1Coding
81.0% best30 / 123
73.9%45 / 123
Legal Research BenchLegal
41.3% best30 / 90
34.1%50 / 90
Tax Agent BenchFinance
34.4% best17 / 82
32.1%23 / 82
Terminal-Bench 4.0Agentic
39.4% best15 / 61
25.3%28 / 61
DeepSWEAgentic
74.9% best8 / 59
73.7%13 / 59
Agent Arena steerabilityAgentic
8.8 best14 / 54
−2.636 / 54
Agent Arena task outcomeAgentic
6.7 best12 / 54
−0.734 / 54
HLAB-Vals-1.1Legal
12.5% best17 / 35
7.5%24 / 35
AutomationBenchAgentic
56.2%4 / 32
62.2% best1 / 32
Agents' Last ExamAgentic
41.0% best8 / 21
29.6%13 / 21