Claude Opus 4.8 vs Gemini 3.7 Flash

Change models
IndexAll categories, out of 100.
66.6 ±2.0
67.2 ±2.2 best
GapIndex points below the top.
−0.6 · no clear gap
Top
CoverageIndex weight that has results.
100%
100%
SpeedOutput tokens per second.
76/s
96/s best
EloLMArena rating and rank.
1454
1489 best
ReleasedMonth of release.
2026-05
2026-08
Input / 1MUS dollars per 1M input tokens.
$5.00
$0.75 best
Output / 1MUS dollars per 1M output tokens.
$25.00
$3.75 best
Batch output / 1MBatch price per 1M output tokens.
$12.50
$1.88 best
1K chats2K tokens in, 500 out, each.
$22.50
$3.38 best
1K documents20K tokens in, 1K out, each.
$125.00
$18.75 best
1K agent tasks200K tokens in, 20K out, each.
$1,500.00
$225.00 best
ContextMaximum tokens in one request.
1M
1.05M best
AgenticMulti-step tasks with tools.
65.2 best
63.6
CodingCode writing and repair.
67.3 best
65.7
ReasoningLogic problems and puzzles.
66.6
67.3 best
MultimodalTasks with images and text.
66.6
67.4 best
KnowledgeFacts and expert knowledge.
65.5
70.1 best
MultilingualTasks in many languages.
59.2
71.9 best
InstructionTasks with strict prompt rules.
58.8
86.0 best
MathMath problems.
69.9 best
65.7

Benchmarks

64 shared
Only Claude Opus 4.8: 40Only Gemini 3.7 Flash: 30
Claude Opus 4.8
Gemini 3.7 Flash
91.0%37 / 220
94.8% best11 / 220
48.7% best15 / 216
47.9%18 / 216
77.7%73 / 211
81.7% best33 / 211
20.9% best25 / 207
14.3%51 / 207
98.3% best31 / 203
97.2%37 / 203
92.0%22 / 193
94.5% best4 / 193
48.8%29 / 170
55.3% best15 / 170
83.6%51 / 149
87.3% best6 / 149
TaxEval v2Finance
75.6% best14 / 143
74.7%31 / 143
87.8%11 / 142
88.7% best5 / 142
72.1%30 / 137
84.6% best17 / 137
92.4%16 / 135
93.9% best5 / 135
MMLU ProKnowledge
89.6%9 / 135
90.1% best6 / 135
92.5%31 / 132
95.5% best20 / 132
47.7% best32 / 132
44.6%40 / 132
82.7% best29 / 130
70.4%55 / 130
54.4%33 / 113
57.2% best21 / 113
MedScribeHealthcare
85.8% best28 / 111
83.9%42 / 111
MedCodeHealthcare
53.2%10 / 108
53.4% best9 / 108
43.8% best28 / 98
34.6%56 / 98