Inference
EN

GPT-5.6 Sol vs Hy4

Change models
GPT-5.6 SolOpenAI
Hy4Tencent
IndexAll categories, out of 100.
73.0 ±1.9 best
72.0 ±7.8
GapIndex points below the top.
Top
−1.0 · no clear gap
CoverageIndex weight that has results.
100% best
65%
SpeedOutput tokens per second.
64/s best
52/s
EloLMArena rating and rank.
1457
—
ReleasedMonth of release.
2026-07
2026-11
Input / 1MUS dollars per 1M input tokens.
$4.40
$0.417 best
Output / 1MUS dollars per 1M output tokens.
$22.00
$1.25 best
Batch output / 1MBatch price per 1M output tokens.
$10.00
—
1K chats2K tokens in, 500 out, each.
$19.80
$1.46 best
1K documents20K tokens in, 1K out, each.
$110.00
$9.59 best
1K agent tasks200K tokens in, 20K out, each.
$1,320.00
$108.41 best
ContextMaximum tokens in one request.
1.05M
1.05M
AgenticMulti-step tasks with tools.
71.9
74.5 best
CodingCode writing and repair.
70.4 best
68.4
ReasoningLogic problems and puzzles.
76.3
—
MultimodalTasks with images and text.
72.6 best
57.4
KnowledgeFacts and expert knowledge.
72.7
83.7 best
MultilingualTasks in many languages.
66.4
—
InstructionTasks with strict prompt rules.
64.5
—
MathMath problems.
75.1
—

Benchmarks

9 shared
Only GPT-5.6 Sol: 116Only Hy4: 12
GPT-5.6 Sol
Hy4
Vibe Code Bench v1.1Coding
80.5%31 / 123
81.0% best30 / 123
Legal Research BenchLegal
48.1% best12 / 90
41.3%30 / 90
Tax Agent BenchFinance
31.1%29 / 82
34.4% best17 / 82
Terminal-Bench 4.0Agentic
37.9%19 / 61
39.4% best15 / 61
DeepSWEAgentic
72.7%15 / 59
74.9% best8 / 59
Agent Arena steerabilityAgentic
5.516 / 54
8.8 best14 / 54
Agent Arena task outcomeAgentic
2.325 / 54
6.7 best12 / 54
Massive Multi-discipline Multimodal Understanding ProMultimodal
83.0% best5 / 53
78.6%22 / 53
FrontierCode 2.0Coding
27.6%12 / 27
31.0% best9 / 27