Artificial Analysis leaderboard
The model pool
Every model you can drop into a tier list, ranked by intelligence index. Pulled straight from the Artificial Analysis API — nothing on this page is hand-curated.
- Models
- 50
- Providers
- 17
- Top score
- 65.7
- Last synced
- Sep 3, 2026
- 1
Claude Fable 5.1
AnthropicSep 2026
Intel
65.7
Coding
81.6
$ / 1M
$20
Tok / s
69
- 2
Claude Opus 5
AnthropicJul 2026
Intel
63.1
Coding
78.0
$ / 1M
$10
Tok / s
48
- 3
Claude Fable 5
AnthropicJun 2026
Intel
62.1
Coding
76.5
$ / 1M
$20
Tok / s
59
- 12
Claude Opus 4.8
AnthropicMay 2026
Intel
57.3
Coding
74.3
$ / 1M
$10
Tok / s
0
- 19
Claude Sonnet 5
AnthropicJun 2026
Intel
55.3
Coding
71.5
$ / 1M
$4.00
Tok / s
78
- 20
Claude Opus 4.7
AnthropicApr 2026
Intel
55.0
Coding
73.6
$ / 1M
$10
Tok / s
0
- 32
Claude Sonnet 4.6
AnthropicFeb 2026
Intel
48.4
Coding
63.0
$ / 1M
$6.00
Tok / s
0
- 41
Claude Opus 4.6
AnthropicFeb 2026
Intel
44.9
Coding
—
$ / 1M
$10
Tok / s
0
Intelligence and coding are Artificial Analysis composite indices. Price is the blended 3:1 input/output cost per million tokens.