Claude Sonnet 4.5 Thinking vs GPT-5

Change models
GPT-5OpenAI
IndexAll categories, out of 100.
53.3 ±6.9
53.5 ±3.4 best
GapIndex points below the top.
−0.2 · no clear gap
Top
CoverageIndex weight that has results.
80%
95% best
SpeedOutput tokens per second.
—
77/s
EloLMArena rating and rank.
—
1406
ReleasedMonth of release.
2025-09
2025-08
Input / 1MUS dollars per 1M input tokens.
—
$1.25
Output / 1MUS dollars per 1M output tokens.
—
$10.00
Batch output / 1MBatch price per 1M output tokens.
—
$5.00
1K chats2K tokens in, 500 out, each.
—
$7.50
1K documents20K tokens in, 1K out, each.
—
$35.00
1K agent tasks200K tokens in, 20K out, each.
—
$450.00
ContextMaximum tokens in one request.
200K
400K best
AgenticMulti-step tasks with tools.
56.6 best
49.8
CodingCode writing and repair.
50.1
51.3 best
ReasoningLogic problems and puzzles.
—
54.3
MultimodalTasks with images and text.
57.7 best
55.9
KnowledgeFacts and expert knowledge.
52.9
57.4 best
MultilingualTasks in many languages.
60.2
—
InstructionTasks with strict prompt rules.
—
56.5
MathMath problems.
53.4
54.8 best

Benchmarks

20 shared
Only Claude Sonnet 4.5 Thinking: 8Only GPT-5: 40
Claude Sonnet 4.5 Thinking
GPT-5
84.1%39 / 149
86.0% best17 / 149
TaxEval v2Finance
73.3%50 / 143
73.4% best49 / 143
73.0%82 / 142
85.9% best25 / 142
81.6%61 / 135
85.6% best45 / 135
MMLU ProKnowledge
87.4% best28 / 135
86.5%41 / 135
CorpFin v2Finance
62.0% best48 / 131
61.1%58 / 131
22.6% best94 / 130
20.1%98 / 130
MedScribeHealthcare
84.1% best40 / 111
83.7%44 / 111
MedCodeHealthcare
44.1%42 / 108
49.6% best21 / 108
64.0%49 / 97
65.5% best41 / 97
AIMEMath
88.2%30 / 95
93.4% best14 / 95
MedQAHealthcare
94.7%15 / 95
96.3% best4 / 95
MMMU ProMultimodal
79.3%48 / 92
81.5% best39 / 92
SWE-benchCoding
70.0% best61 / 87
69.0%67 / 87
MGSMMultilingual
94.3% best4 / 74
92.8%14 / 74
41.6% best29 / 66
37.1%38 / 66
IOI v1Coding
18.3%25 / 61
20.0% best24 / 61
62.2%19 / 53
66.5% best6 / 53
54.5% best16 / 51
52.2%25 / 51
61.3% best2 / 47
48.8%13 / 47