Claude 3.5 Sonnet
LatestAnthropic
•
Proprietary# 110
Released
Oct 22, 2024
# 16
Knowledge Cutoff
Apr 24
# 9
Context Length
200K
Benchmarks
# 138
Code RankedAGI
41.0%
# 52
SWEBench Verified
49.0%
# 160
Agentic RankedAGI
32.5%
# 40
LiveCodeBench v6
36.4%
# 30
LiveCodeBench v5
39.8%
# 18
Code LMArena
1313
# 31
Codeforces ELO
717
# 27
Aider Polyglot
51.6%
# 10
Code LiveBench (old)
67.1%
# 174
Reason RankedAGI
35.9%
# 75
HLE
4.8%
# 79
GPQA Diamond
65.0%
# 59
Text Arena
1355
# 66
AIME 2025 I & II
3.0%
# 44
AIME 2024
16.0%
# 1
Human Eval
93.7%
# 6
Human Eval+
86.2%
# 28
NYT Connections
17.7%
# 27
MMLU Pro
78.0%
# 16
MMLU
88.0%
# 26
MMMU
70.4%
# 28
Halluc. Hughes
4.6%
# 3
Aidan Bench
2691
# 20
Avg LiveBench (old)
60.7%
# 7
IF Evaluation
89.3%
# 35
Coding LiveBench 25.4
32.3%
# 31
Data LiveBench
52.8%
# 9
Language LiveBench
53.8%
# 5
Quality Artificial Analysis
80
# 211
Math RankedAGI
37.4%
# 156
RAGI RankedAGI
40.9%
Pricing
# 39
Input Cost /M
$3
# 46
Output Cost /M
$15
# 26
Cached Cost /M
$0.3