Claude Opus 5 vs GPT-6.2 Astra

Change models
Claude Opus 5Anthropic
IndexAll categories, out of 100.
72.1 ±1.9
79.5 ±3.1 best
GapIndex points below the top.
−7.4
Top
CoverageIndex weight that has results.
100% best
95%
SpeedOutput tokens per second.
68/s best
39/s
EloLMArena rating and rank.
1502 best
1414
ReleasedMonth of release.
2026-07
2026-11
Input / 1MUS dollars per 1M input tokens.
$5.00 best
$8.00
Output / 1MUS dollars per 1M output tokens.
$25.00 best
$40.00
Batch output / 1MBatch price per 1M output tokens.
$12.50 best
$20.00
1K chats2K tokens in, 500 out, each.
$22.50 best
$36.00
1K documents20K tokens in, 1K out, each.
$125.00 best
$200.00
1K agent tasks200K tokens in, 20K out, each.
$1,500.00 best
$2,400.00
ContextMaximum tokens in one request.
1M
1.05M best
AgenticMulti-step tasks with tools.
70.1
74.7 best
CodingCode writing and repair.
73.4
79.4 best
ReasoningLogic problems and puzzles.
73.5
86.2 best
MultimodalTasks with images and text.
67.1
70.7 best
KnowledgeFacts and expert knowledge.
73.0
79.9 best
MultilingualTasks in many languages.
62.3
74.0 best
InstructionTasks with strict prompt rules.
55.1
—
MathMath problems.
73.3
76.4 best

Benchmarks

37 shared
Only Claude Opus 5: 114Only GPT-6.2 Astra: 26
Claude Opus 5
GPT-6.2 Astra
93.9%18 / 220
96.8% best1 / 220
98.9%25 / 203
100.0% best1 / 203
60.9%9 / 170
68.6% best2 / 170
90.4%10 / 137
95.8% best1 / 137
97.5%7 / 132
98.5% best1 / 132
88.4%13 / 130
89.9% best9 / 130
97.1%8 / 127
97.5% best6 / 127
99.0%3 / 116
99.3% best2 / 116
84.7%7 / 109
89.5% best1 / 109
55.3% best3 / 98
42.8%31 / 98
59.9%22 / 94
78.5% best2 / 94
Chess PuzzlesReasoning
43.0%34 / 93
78.0% best3 / 93
85.6%18 / 93
92.6% best4 / 93
47.1% best7 / 89
21.8%54 / 89
59.0%12 / 83
95.0% best1 / 83
73.2%18 / 75
100.0% best1 / 75
53.5%14 / 68
63.1% best4 / 68
2.422 / 59
14.0 best7 / 59
10.114 / 59
12.4 best10 / 59
12.0%13 / 55
17.0% best4 / 55