Claude 3.5 Sonnet
LatestAnthropic
•
Proprietary# 47
Released
Oct 22, 2024
# 9
Knowledge Cutoff
Apr 24
# 6
Context Length
200K
Benchmarks
# 39
Code RankedAGI
48.3%
# 30
Code LiveBench
32.3%
# 14
Aider Polyglot
51.6%
# 10
SWEBench Verified
49.0%
# 23
Codeforces ELO
717
# 5
WebDev Arena
1245.4
# 27
LiveCodeBench v5
39.8%
# 13
Code LMArena
1313
# 10
Code LiveBench 24.11
67.1%
# 26
GPQA Diamond
65.0%
# 20
Reason LiveBench
58.7%
# 29
ELO LMArena
1282
# 31
AIME 2025 I & II
3.0%
# 27
Math LiveBench
51.3%
# 17
MATH
78.3%
# 22
MATH 500
78.3%
# 1
Human Eval
93.7%
# 6
Human Eval+
86.2%
# 19
NYT Connections
17.7%
# 10
MMLU Pro
78.0%
# 10
MMLU
88.0%
# 14
MMMU
70.4%
# 22
Halluc. Hughes
4.6%
# 3
Aidan Bench
2691
# 40
AIME 2024
16.0%
# 31
IF LiveBench
69.3%
# 31
Data LiveBench
52.8%
# 9
Lang LiveBench
53.8%
# 20
Avg LiveBench
60.7%
# 5
IF Evaluation
89.3%
Pricing
# 27
Input Cost /M
$3
# 33
Output Cost /M
$15
# 10
Cached Cost /M
$0.3