Inference
EN

Claude Sonnet 5.5 vs GPT-6.1 Terra

Change models
Claude Sonnet 5.5Anthropic
GPT-6.1 TerraOpenAI
IndexAll categories, out of 100.
77.2 ±2.6 best
77.1 ±3.8
GapIndex points below the top.
Top
−0.1 · no clear gap
CoverageIndex weight that has results.
95% best
90%
SpeedOutput tokens per second.
107/s best
46/s
EloLMArena rating and rank.
1469
—
ReleasedMonth of release.
2026-09
2026-10
Input / 1MUS dollars per 1M input tokens.
$2.00
$1.00 best
Output / 1MUS dollars per 1M output tokens.
$10.00
$5.00 best
Batch output / 1MBatch price per 1M output tokens.
$5.00
$2.50 best
1K chats2K tokens in, 500 out, each.
$9.00
$4.50 best
1K documents20K tokens in, 1K out, each.
$50.00
$25.00 best
1K agent tasks200K tokens in, 20K out, each.
$600.00
$300.00 best
ContextMaximum tokens in one request.
1M
1.05M best
AgenticMulti-step tasks with tools.
77.5
78.6 best
CodingCode writing and repair.
80.8 best
74.4
ReasoningLogic problems and puzzles.
77.8
81.3 best
MultimodalTasks with images and text.
74.4 best
73.2
KnowledgeFacts and expert knowledge.
71.3 best
70.5
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
41.5
—
MathMath problems.
76.1
76.9 best

Benchmarks

45 shared
Only Claude Sonnet 5.5: 83Only GPT-6.1 Terra: 5
Claude Sonnet 5.5
GPT-6.1 Terra
Artificial Analysis Humanity's Last ExamKnowledge
55.0% best6 / 222
46.2%24 / 222
Artificial Analysis Long Context ReasoningReasoning
82.7%23 / 217
82.7%23 / 217
GPQA diamond · Epoch AIKnowledge
95.6% best4 / 217
94.3%10 / 217
Critical Physics TasksReasoning
31.4% best7 / 214
28.0%16 / 214
OTIS Mock AIME 2024-2025Math
100.0%1 / 200
100.0%1 / 200
Artificial Analysis Omniscience AccuracyKnowledge
51.5% best17 / 164
47.6%28 / 164
Vibe Code Bench v1.1Coding
92.4% best2 / 123
84.0%19 / 123
Artificial Analysis SciCodeCoding
61.0% best6 / 114
55.1%28 / 114
SimpleQA VerifiedKnowledge
46.5%39 / 91
57.3% best21 / 91
FrontierMath-Tiers-1-3-v2-PrivateMath
88.8%10 / 90
90.9% best5 / 90
Legal Research BenchLegal
48.1% best12 / 90
26.0%62 / 90
Tax Agent BenchFinance
42.0% best10 / 82
17.1%58 / 82
LiveBench Data AnalysisReasoning
59.5%73 / 80
82.3% best4 / 80
LiveBench ReasoningReasoning
91.6% best7 / 80
88.7%20 / 80
Mystery Game PuzzlesReasoning
65.0%7 / 80
74.0% best4 / 80
FrontierMath-Tier-4-v2-PrivateMath
80.5%11 / 72
92.7% best6 / 72
Terminal-Bench 4.0Agentic
64.1% best3 / 61
49.5%11 / 61
Agent Arena steerabilityAgentic
14.57 / 54
16.6 best4 / 54
Agent Arena task outcomeAgentic
19.5 best3 / 54
9.711 / 54
LiveBench Coding 2.0Coding
94.7% best6 / 52
90.8%11 / 52