Rankings — WorldAttention

All models (49)

1
Gemini 4 Argon
Closed
Google
Agentic
87.4
Data completeness
65%
Input / output per 1M (USD)
$2 / $10.0
2026-09-30
View details
Google · 2026-09-30
Current rank
#1
Data sources
3
Capabilities and license
Reasoning
Not measured / insufficient evidence
Coding
Not measured / insufficient evidence
Mathematics
Not measured / insufficient evidence
Agentic
87.40
Mercor APEX · 2026-10-18
Human preference
1523.90
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
2
Claude Sonnet 5.5
Closed
Anthropic
Agentic
83.7
Data completeness
85%
Input / output per 1M (USD)
$2 / $10.0
2026-09-28
View details
Anthropic · 2026-09-28
Current rank
#2
Data sources
4
Capabilities and license
Reasoning
91.63
LiveBench · Reasoning · 2026-10-18
Coding
91.37
LiveBench · Coding · 2026-10-18
Mathematics
96.69
LiveBench · Mathematics · 2026-10-18
Agentic
83.70
Mercor APEX · 2026-10-18
Human preference
1474.94
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
3
Claude Opus 5.5
Closed
Anthropic
Agentic
81.3
Data completeness
85%
Input / output per 1M (USD)
$4 / $20.0
2026-09-22
View details
Anthropic · 2026-09-22
Current rank
#3
Data sources
4
Capabilities and license
Reasoning
92.15
LiveBench · Reasoning · 2026-10-18
Coding
89.25
LiveBench · Coding · 2026-10-18
Mathematics
97.08
LiveBench · Mathematics · 2026-10-18
Agentic
81.30
Mercor APEX · 2026-10-18
Human preference
1508.17
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
4
Gemini 3.7 Flash
Closed
Google
Agentic
79.4
Data completeness
100%
Input / output per 1M (USD)
$0.75 / $3.75
2026-08-13
View details
Google · 2026-08-13
Current rank
#4
Data sources
5
Capabilities and license
Reasoning
87.80
LiveBench · Reasoning · 2026-10-18
Coding
78.89
LiveBench · Coding · 2026-10-18
Mathematics
93.47
LiveBench · Mathematics · 2026-10-18
Agentic
79.40
Mercor APEX · 2026-10-18
Human preference
1487.59
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
5
Grok 4.6
Closed
SpaceXAI
Agentic
77.3
Data completeness
100%
Input / output per 1M (USD)
$2 / $6
2026-08-12
View details
SpaceXAI · 2026-08-12
Current rank
#5
Data sources
5
Capabilities and license
Reasoning
90.51
LiveBench · Reasoning · 2026-10-18
Coding
76.78
LiveBench · Coding · 2026-10-18
Mathematics
92.57
LiveBench · Mathematics · 2026-10-18
Agentic
77.30
Mercor APEX · 2026-10-18
Human preference
1454.49
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
6
Claude Fable 5.1
Closed
Anthropic
Agentic
77.1
Data completeness
85%
Input / output per 1M (USD)
$10.0 / $50.0
2026-09-01
View details
Anthropic · 2026-09-01
Current rank
#6
Data sources
4
Capabilities and license
Reasoning
91.69
LiveBench · Reasoning · 2026-10-18
Coding
86.38
LiveBench · Coding · 2026-10-18
Mathematics
97.01
LiveBench · Mathematics · 2026-10-18
Agentic
77.10
Mercor APEX · 2026-10-18
Human preference
1502.29
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
7
Claude Opus 5
Closed
Anthropic
Agentic
77
Data completeness
100%
Input / output per 1M (USD)
$5 / $25.0
2026-07-24
View details
Anthropic · 2026-07-24
Current rank
#7
Data sources
5
Capabilities and license
Reasoning
91.21
LiveBench · Reasoning · 2026-10-18
Coding
81.45
LiveBench · Coding · 2026-10-18
Mathematics
95.73
LiveBench · Mathematics · 2026-10-18
Agentic
77.00
Mercor APEX · 2026-10-18
Human preference
1489.99
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
8
GPT-6 Astra
Unverified
OpenAI
Agentic
75.1
Data completeness
80%
Input / output per 1M (USD)
$10.0 / $50.0
2026-09-03
View details
OpenAI · 2026-09-03
Current rank
#8
Data sources
4
Capabilities and license
Reasoning
92.65
LiveBench · Reasoning · 2026-10-18
Coding
80.36
LiveBench · Coding · 2026-10-18
Mathematics
96.81
LiveBench · Mathematics · 2026-10-18
Agentic
75.10
Mercor APEX · 2026-10-18
Human preference
Not measured / insufficient evidence
Unverified
9
Claude Fable 5
Closed
Anthropic
Agentic
74.5
Data completeness
100%
Input / output per 1M (USD)
$10.0 / $50.0
2026-06-09
View details
Anthropic · 2026-06-09
Current rank
#9
Data sources
5
Capabilities and license
Reasoning
89.65
LiveBench · Reasoning · 2026-10-18
Coding
85.99
LiveBench · Coding · 2026-10-18
Mathematics
95.99
LiveBench · Mathematics · 2026-10-18
Agentic
74.50
Mercor APEX · 2026-10-18
Human preference
1503.60
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
10
Qwen3.8 Max
Closed
Alibaba
Agentic
74.3
Data completeness
100%
Input / output per 1M (USD)
$2 / $6
2026-09-02
View details
Alibaba · 2026-09-02
Current rank
#10
Data sources
5
Capabilities and license
Reasoning
88.21
LiveBench · Reasoning · 2026-10-18
Coding
72.87
LiveBench · Coding · 2026-10-18
Mathematics
91.31
LiveBench · Mathematics · 2026-10-18
Agentic
74.30
Mercor APEX · 2026-10-18
Human preference
1482.35
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
11
GPT-6.1 Sol
Unverified
OpenAI
Agentic
73.2
Data completeness
65%
Input / output per 1M (USD)
$2 / $10.0
2026-09-29
View details
OpenAI · 2026-09-29
Current rank
#11
Data sources
3
Capabilities and license
Reasoning
91.63
LiveBench · Reasoning · 2026-10-18
Coding
80.74
LiveBench · Coding · 2026-10-18
Mathematics
96.48
LiveBench · Mathematics · 2026-10-18
Agentic
73.20
Mercor APEX · 2026-10-18
Human preference
Not measured / insufficient evidence
Unverified
12
Gemini 3.8 Flash
Closed
Google
Agentic
72.9
Data completeness
100%
Input / output per 1M (USD)
$0.75 / $3.75
2026-09-02
View details
Google · 2026-09-02
Current rank
#12
Data sources
5
Capabilities and license
Reasoning
89.29
LiveBench · Reasoning · 2026-10-18
Coding
72.49
LiveBench · Coding · 2026-10-18
Mathematics
91.56
LiveBench · Mathematics · 2026-10-18
Agentic
72.90
Mercor APEX · 2026-10-18
Human preference
1496.38
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
13
GPT-5.6 Terra
Closed
OpenAI
Agentic
71.8
Data completeness
100%
Input / output per 1M (USD)
$2 / $12.0
2026-07-09
View details
OpenAI · 2026-07-09
Current rank
#13
Data sources
5
Capabilities and license
Reasoning
90.63
LiveBench · Reasoning · 2026-10-18
Coding
78.25
LiveBench · Coding · 2026-10-18
Mathematics
94.91
LiveBench · Mathematics · 2026-10-18
Agentic
71.80
Mercor APEX · 2026-10-18
Human preference
1466.43
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
14
Grok 4.5
Closed
SpaceXAI
Agentic
70.5
Data completeness
100%
Input / output per 1M (USD)
$2 / $6
2026-07-08
View details
SpaceXAI · 2026-07-08
Current rank
#14
Data sources
5
Capabilities and license
Reasoning
87.17
LiveBench · Reasoning · 2026-10-18
Coding
68.59
LiveBench · Coding · 2026-10-18
Mathematics
90.82
LiveBench · Mathematics · 2026-10-18
Agentic
70.50
Mercor APEX · 2026-10-18
Human preference
1465.71
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
15
Muse Spark 1.3
Closed
Meta
Agentic
70.4
Data completeness
85%
Input / output per 1M (USD)
$1.25 / $4.25
2026-09-02
View details
Meta · 2026-09-02
Current rank
#15
Data sources
4
Capabilities and license
Reasoning
89.65
LiveBench · Reasoning · 2026-10-18
Coding
81.06
LiveBench · Coding · 2026-10-18
Mathematics
95.95
LiveBench · Mathematics · 2026-10-18
Agentic
70.40
Mercor APEX · 2026-10-18
Human preference
1494.41
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
16
GPT-5.5
Closed
OpenAI
Agentic
70.3
Data completeness
100%
Input / output per 1M (USD)
$5 / $30.0
2026-04-23
View details
OpenAI · 2026-04-23
Current rank
#16
Data sources
5
Capabilities and license
Reasoning
89.65
LiveBench · Reasoning · 2026-10-18
Coding
82.15
LiveBench · Coding · 2026-10-18
Mathematics
95.86
LiveBench · Mathematics · 2026-10-18
Agentic
70.30
Mercor APEX · 2026-10-18
Human preference
1481.44
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
17
GPT-6 Sol
Unverified
OpenAI
Agentic
68.7
Data completeness
45%
Input / output per 1M (USD)
$2 / $10.0
2026-09-22
View details
OpenAI · 2026-09-22
Current rank
#17
Data sources
2
Capabilities and license
Reasoning
Not measured / insufficient evidence
Coding
Not measured / insufficient evidence
Mathematics
Not measured / insufficient evidence
Agentic
68.70
Mercor APEX · 2026-10-18
Human preference
Not measured / insufficient evidence
Unverified
18
Grok 4.7
Closed
SpaceXAI
Agentic
68.7
Data completeness
85%
Input / output per 1M (USD)
$2 / $6
2026-09-21
View details
SpaceXAI · 2026-09-21
Current rank
#18
Data sources
4
Capabilities and license
Reasoning
82.65
LiveBench · Reasoning · 2026-10-18
Coding
77.16
LiveBench · Coding · 2026-10-18
Mathematics
95.68
LiveBench · Mathematics · 2026-10-18
Agentic
68.70
Mercor APEX · 2026-10-18
Human preference
1443.98
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
19
Claude Haiku 5.5
Unverified
Anthropic
Agentic
68.3
Data completeness
65%
Input / output per 1M (USD)
$0.1 / $0.5
2026-10-07
View details
Anthropic · 2026-10-07
Current rank
#19
Data sources
3
Capabilities and license
Reasoning
84.63
LiveBench · Reasoning · 2026-10-18
Coding
77.86
LiveBench · Coding · 2026-10-18
Mathematics
92.75
LiveBench · Mathematics · 2026-10-18
Agentic
68.30
Mercor APEX · 2026-10-18
Human preference
Not measured / insufficient evidence
Unverified
20
GPT-5.4
Closed
OpenAI
Agentic
68.2
Data completeness
100%
Input / output per 1M (USD)
$2.5 / $15.0
2026-03-05
View details
OpenAI · 2026-03-05
Current rank
#20
Data sources
5
Capabilities and license
Reasoning
88.12
LiveBench · Reasoning · 2026-10-18
Coding
77.54
LiveBench · Coding · 2026-10-18
Mathematics
94.15
LiveBench · Mathematics · 2026-10-18
Agentic
68.20
Mercor APEX · 2026-10-18
Human preference
1475.12
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
21
DeepSeek V4 Flash
Open license
DeepSeek
Agentic
67.5
Data completeness
100%
Input / output per 1M (USD)
$0.14 / $0.28
2026-07-31
View details
DeepSeek · 2026-07-31
Current rank
#21
Data sources
5
Capabilities and license
Reasoning
86.63
LiveBench · Reasoning · 2026-10-18
Coding
74.98
LiveBench · Coding · 2026-10-18
Mathematics
86.79
LiveBench · Mathematics · 2026-10-18
Agentic
67.50
Mercor APEX · 2026-10-18
Human preference
1438.67
Arena · Text style control (Elo) · 2026-10-18
MIT · Arena
22
Claude Sonnet 5
Closed
Anthropic
Agentic
67.4
Data completeness
100%
Input / output per 1M (USD)
$2 / $10.0
2026-06-30
View details
Anthropic · 2026-06-30
Current rank
#22
Data sources
5
Capabilities and license
Reasoning
88.69
LiveBench · Reasoning · 2026-10-18
Coding
80.68
LiveBench · Coding · 2026-10-18
Mathematics
92.94
LiveBench · Mathematics · 2026-10-18
Agentic
67.40
Mercor APEX · 2026-10-18
Human preference
1461.07
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
23
GLM-5.3-Flash
Open license
Z AI
Agentic
66.8
Data completeness
100%
Input / output per 1M (USD)
$0.15 / $0.5
2026-08-26
View details
Z AI · 2026-08-26
Current rank
#23
Data sources
5
Capabilities and license
Reasoning
77.64
LiveBench · Reasoning · 2026-10-18
Coding
78.95
LiveBench · Coding · 2026-10-18
Mathematics
81.24
LiveBench · Mathematics · 2026-10-18
Agentic
66.80
Mercor APEX · 2026-10-18
Human preference
1473.84
Arena · Text style control (Elo) · 2026-10-18
MIT · Arena
24
GLM-5.3
Open license
Z AI
Agentic
66.2
Data completeness
100%
Input / output per 1M (USD)
$1.4 / $4.4
2026-08-18
View details
Z AI · 2026-08-18
Current rank
#24
Data sources
5
Capabilities and license
Reasoning
85.80
LiveBench · Reasoning · 2026-10-18
Coding
78.95
LiveBench · Coding · 2026-10-18
Mathematics
87.90
LiveBench · Mathematics · 2026-10-18
Agentic
66.20
Mercor APEX · 2026-10-18
Human preference
1478.54
Arena · Text style control (Elo) · 2026-10-18
MIT · Arena
25
MiMo-V2.6-Pro
Open license
Xiaomi
Agentic
65.3
Data completeness
65%
Input / output per 1M (USD)
$0.43 / $0.87
2026-09-21
View details
Xiaomi · 2026-09-21
Current rank
#25
Data sources
3
Capabilities and license
Reasoning
Not measured / insufficient evidence
Coding
Not measured / insufficient evidence
Mathematics
Not measured / insufficient evidence
Agentic
65.30
Mercor APEX · 2026-10-18
Human preference
1479.19
Arena · Text style control (Elo) · 2026-10-18
MIT · Arena
26
GPT-5.6 Sol
Closed
OpenAI
Agentic
65.2
Data completeness
100%
Input / output per 1M (USD)
$4 / $20.0
2026-07-09
View details
OpenAI · 2026-07-09
Current rank
#26
Data sources
5
Capabilities and license
Reasoning
91.65
LiveBench · Reasoning · 2026-10-18
Coding
83.94
LiveBench · Coding · 2026-10-18
Mathematics
96.20
LiveBench · Mathematics · 2026-10-18
Agentic
65.20
Mercor APEX · 2026-10-18
Human preference
1484.74
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
27
Claude Opus 4.8
Closed
Anthropic
Agentic
64.4
Data completeness
100%
Input / output per 1M (USD)
$5 / $25.0
2026-05-28
View details
Anthropic · 2026-05-28
Current rank
#27
Data sources
5
Capabilities and license
Reasoning
89.19
LiveBench · Reasoning · 2026-10-18
Coding
81.83
LiveBench · Coding · 2026-10-18
Mathematics
94.32
LiveBench · Mathematics · 2026-10-18
Agentic
64.40
Mercor APEX · 2026-10-18
Human preference
1481.50
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
28
Gemini 3.6 Flash
Closed
Google
Agentic
63.4
Data completeness
100%
Input / output per 1M (USD)
$0.75 / $3.75
2026-07-21
View details
Google · 2026-07-21
Current rank
#28
Data sources
5
Capabilities and license
Reasoning
85.15
LiveBench · Reasoning · 2026-10-18
Coding
77.86
LiveBench · Coding · 2026-10-18
Mathematics
86.40
LiveBench · Mathematics · 2026-10-18
Agentic
63.40
Mercor APEX · 2026-10-18
Human preference
1482.70
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
29
Claude Opus 4.7
Closed
Anthropic
Agentic
63.1
Data completeness
85%
Input / output per 1M (USD)
$5 / $25.0
2026-04-16
View details
Anthropic · 2026-04-16
Current rank
#29
Data sources
4
Capabilities and license
Reasoning
87.19
LiveBench · Reasoning · 2026-10-18
Coding
82.09
LiveBench · Coding · 2026-10-18
Mathematics
92.85
LiveBench · Mathematics · 2026-10-18
Agentic
63.10
Mercor APEX · 2026-10-18
Human preference
1501.06
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
30
DeepSeek V4 Pro
Open license
DeepSeek
Agentic
62.3
Data completeness
100%
Input / output per 1M (USD)
$1.32 / $3.96
2026-08-13
View details
DeepSeek · 2026-08-13
Current rank
#30
Data sources
5
Capabilities and license
Reasoning
85.84
LiveBench · Reasoning · 2026-10-18
Coding
77.16
LiveBench · Coding · 2026-10-18
Mathematics
95.09
LiveBench · Mathematics · 2026-10-18
Agentic
62.30
Mercor APEX · 2026-10-18
Human preference
1464.58
Arena · Text style control (Elo) · 2026-10-18
MIT · Arena
31
Qwen3.8 27B
Open license
Alibaba
Agentic
62.2
Data completeness
85%
Input / output per 1M (USD)
$0.5 / $3
2026-08-14
View details
Alibaba · 2026-08-14
Current rank
#31
Data sources
4
Capabilities and license
Reasoning
80.03
LiveBench · Reasoning · 2026-10-18
Coding
75.69
LiveBench · Coding · 2026-10-18
Mathematics
86.21
LiveBench · Mathematics · 2026-10-18
Agentic
62.20
Mercor APEX · 2026-10-18
Human preference
1437.75
Arena · Text style control (Elo) · 2026-10-18
Apache 2.0 · Arena
32
Kimi K3
Open weights
Kimi
Agentic
62.2
Data completeness
100%
Input / output per 1M (USD)
$3 / $15.0
2026-07-16
View details
Kimi · 2026-07-16
Current rank
#32
Data sources
5
Capabilities and license
Reasoning
90.67
LiveBench · Reasoning · 2026-10-18
Coding
81.45
LiveBench · Coding · 2026-10-18
Mathematics
84.44
LiveBench · Mathematics · 2026-10-18
Agentic
62.20
Mercor APEX · 2026-10-18
Human preference
1487.07
Arena · Text style control (Elo) · 2026-10-18
Kimi K3 license · Arena
33
Claude Opus 4.6
Closed
Anthropic
Agentic
61.8
Data completeness
85%
Input / output per 1M (USD)
$5 / $25.0
2026-02-05
View details
Anthropic · 2026-02-05
Current rank
#33
Data sources
4
Capabilities and license
Reasoning
88.67
LiveBench · Reasoning · 2026-10-18
Coding
78.18
LiveBench · Coding · 2026-10-18
Mathematics
89.32
LiveBench · Mathematics · 2026-10-18
Agentic
61.80
Mercor APEX · 2026-10-18
Human preference
1503.05
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
34
GLM-5.2
Open license
Z AI
Agentic
61.1
Data completeness
100%
Input / output per 1M (USD)
$1.4 / $4.4
2026-06-16
View details
Z AI · 2026-06-16
Current rank
#34
Data sources
5
Capabilities and license
Reasoning
78.63
LiveBench · Reasoning · 2026-10-18
Coding
79.65
LiveBench · Coding · 2026-10-18
Mathematics
89.78
LiveBench · Mathematics · 2026-10-18
Agentic
61.10
Mercor APEX · 2026-10-18
Human preference
1475.88
Arena · Text style control (Elo) · 2026-10-18
MIT · Arena
35
Muse Spark 1.2
Closed
Meta
Agentic
59.8
Data completeness
100%
Input / output per 1M (USD)
$1.25 / $4.25
2026-08-05
View details
Meta · 2026-08-05
Current rank
#35
Data sources
5
Capabilities and license
Reasoning
90.00
LiveBench · Reasoning · 2026-10-18
Coding
77.54
LiveBench · Coding · 2026-10-18
Mathematics
91.20
LiveBench · Mathematics · 2026-10-18
Agentic
59.80
Mercor APEX · 2026-10-18
Human preference
1493.06
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
36
GPT-5.6 Luna
Closed
OpenAI
Agentic
57.8
Data completeness
100%
Input / output per 1M (USD)
$0.2 / $1.2
2026-07-09
View details
OpenAI · 2026-07-09
Current rank
#36
Data sources
5
Capabilities and license
Reasoning
85.64
LiveBench · Reasoning · 2026-10-18
Coding
82.92
LiveBench · Coding · 2026-10-18
Mathematics
87.20
LiveBench · Mathematics · 2026-10-18
Agentic
57.80
Mercor APEX · 2026-10-18
Human preference
1452.57
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
37
Claude Sonnet 4.6
Closed
Anthropic
Agentic
57.6
Data completeness
100%
Input / output per 1M (USD)
$3 / $15.0
2026-02-17
View details
Anthropic · 2026-02-17
Current rank
#37
Data sources
5
Capabilities and license
Reasoning
84.77
LiveBench · Reasoning · 2026-10-18
Coding
79.27
LiveBench · Coding · 2026-10-18
Mathematics
86.99
LiveBench · Mathematics · 2026-10-18
Agentic
57.60
Mercor APEX · 2026-10-18
Human preference
1472.10
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
38
GLM-5.1
Open license
Z AI
Agentic
56.8
Data completeness
65%
Input / output per 1M (USD)
$1.28 / $4.07
2026-04-07
View details
Z AI · 2026-04-07
Current rank
#38
Data sources
3
Capabilities and license
Reasoning
Not measured / insufficient evidence
Coding
Not measured / insufficient evidence
Mathematics
Not measured / insufficient evidence
Agentic
56.80
Mercor APEX · 2026-10-18
Human preference
1465.17
Arena · Text style control (Elo) · 2026-10-18
MIT · Arena
39
MiniMax-M3
Open weights
MiniMax
Agentic
53.5
Data completeness
85%
Input / output per 1M (USD)
$0.3 / $1.2
2026-06-01
View details
MiniMax · 2026-06-01
Current rank
#39
Data sources
4
Capabilities and license
Reasoning
74.48
LiveBench · Reasoning · 2026-10-18
Coding
68.20
LiveBench · Coding · 2026-10-18
Mathematics
76.95
LiveBench · Mathematics · 2026-10-18
Agentic
53.50
Mercor APEX · 2026-10-18
Human preference
1441.32
Arena · Text style control (Elo) · 2026-10-18
MiniMax Community License · Arena
40
Kimi K2.7 Code
Unverified
Kimi
Agentic
53.4
Data completeness
80%
Input / output per 1M (USD)
$0.95 / $4
2026-06-12
View details
Kimi · 2026-06-12
Current rank
#40
Data sources
4
Capabilities and license
Reasoning
82.81
LiveBench · Reasoning · 2026-10-18
Coding
73.96
LiveBench · Coding · 2026-10-18
Mathematics
79.60
LiveBench · Mathematics · 2026-10-18
Agentic
53.40
Mercor APEX · 2026-10-18
Human preference
Not measured / insufficient evidence
Unverified
41
Gemini 3.1 Pro
Closed
Google
Agentic
52.2
Data completeness
100%
Input / output per 1M (USD)
$2 / $12.0
2026-02-19
View details
Google · 2026-02-19
Current rank
#41
Data sources
5
Capabilities and license
Reasoning
84.00
LiveBench · Reasoning · 2026-10-18
Coding
76.45
LiveBench · Coding · 2026-10-18
Mathematics
91.05
LiveBench · Mathematics · 2026-10-18
Agentic
52.20
Mercor APEX · 2026-10-18
Human preference
1486.91
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
42
DeepSeek V4.1 Flash
Open license
DeepSeek
Agentic
50.5
Data completeness
85%
Input / output per 1M (USD)
$0.3 / $1.2
2026-09-10
View details
DeepSeek · 2026-09-10
Current rank
#42
Data sources
4
Capabilities and license
Reasoning
86.69
LiveBench · Reasoning · 2026-10-18
Coding
80.04
LiveBench · Coding · 2026-10-18
Mathematics
93.29
LiveBench · Mathematics · 2026-10-18
Agentic
50.50
Mercor APEX · 2026-10-18
Human preference
1474.63
Arena · Text style control (Elo) · 2026-10-18
MIT · Arena
43
Inkling
Open license
Thinking Machines
Agentic
48.8
Data completeness
85%
Input / output per 1M (USD)
$1 / $4.05
2026-07-15
View details
Thinking Machines · 2026-07-15
Current rank
#43
Data sources
4
Capabilities and license
Reasoning
78.35
LiveBench · Reasoning · 2026-10-18
Coding
71.02
LiveBench · Coding · 2026-10-18
Mathematics
88.36
LiveBench · Mathematics · 2026-10-18
Agentic
48.80
Mercor APEX · 2026-10-18
Human preference
1440.87
Arena · Text style control (Elo) · 2026-10-18
Apache 2.0 · Arena
44
Muse Spark 1.1
Closed
Meta
Agentic
46.2
Data completeness
100%
Input / output per 1M (USD)
$1.25 / $4.25
2026-07-09
View details
Meta · 2026-07-09
Current rank
#44
Data sources
5
Capabilities and license
Reasoning
87.73
LiveBench · Reasoning · 2026-10-18
Coding
77.16
LiveBench · Coding · 2026-10-18
Mathematics
87.14
LiveBench · Mathematics · 2026-10-18
Agentic
46.20
Mercor APEX · 2026-10-18
Human preference
1490.18
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
45
Gemini 3.5 Flash-Lite
Closed
Google
Agentic
46.1
Data completeness
85%
Input / output per 1M (USD)
$0.3 / $2.5
2026-07-21
View details
Google · 2026-07-21
Current rank
#45
Data sources
4
Capabilities and license
Reasoning
60.19
LiveBench · Reasoning · 2026-10-18
Coding
76.07
LiveBench · Coding · 2026-10-18
Mathematics
73.74
LiveBench · Mathematics · 2026-10-18
Agentic
46.10
Mercor APEX · 2026-10-18
Human preference
1454.76
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
46
Gemini 3.5 Flash
Closed
Google
Agentic
44.3
Data completeness
100%
Input / output per 1M (USD)
$1.5 / $9
2026-05-19
View details
Google · 2026-05-19
Current rank
#46
Data sources
5
Capabilities and license
Reasoning
82.00
LiveBench · Reasoning · 2026-10-18
Coding
78.18
LiveBench · Coding · 2026-10-18
Mathematics
88.24
LiveBench · Mathematics · 2026-10-18
Agentic
44.30
Mercor APEX · 2026-10-18
Human preference
1477.47
Arena · Text style control (Elo) · 2026-10-18
Proprietary · Arena
47
Qwen3.5 397B A17B
Open license
Alibaba
Agentic
40.3
Data completeness
65%
Input / output per 1M (USD)
$0.6 / $3.6
2026-02-16
View details
Alibaba · 2026-02-16
Current rank
#47
Data sources
3
Capabilities and license
Reasoning
Not measured / insufficient evidence
Coding
Not measured / insufficient evidence
Mathematics
Not measured / insufficient evidence
Agentic
40.30
Mercor APEX · 2026-10-18
Human preference
1441.57
Arena · Text style control (Elo) · 2026-10-18
Apache 2.0 · Arena
48
DeepSeek V3.2
Open license
DeepSeek
Agentic
35.1
Data completeness
65%
Input / output per 1M (USD)
$0.28 / $0.42
2025-12-01
View details
DeepSeek · 2025-12-01
Current rank
#48
Data sources
3
Capabilities and license
Reasoning
Not measured / insufficient evidence
Coding
Not measured / insufficient evidence
Mathematics
Not measured / insufficient evidence
Agentic
35.10
Mercor APEX · 2026-10-18
Human preference
1425.09
Arena · Text style control (Elo) · 2026-10-18
MIT · Arena
49
gpt-oss-120b
Open license
OpenAI
Agentic
8.3
Data completeness
65%
Input / output per 1M (USD)
$0.13 / $0.59
2025-08-05
View details
OpenAI · 2025-08-05
Current rank
#49
Data sources
3
Capabilities and license
Reasoning
Not measured / insufficient evidence
Coding
Not measured / insufficient evidence
Mathematics
Not measured / insufficient evidence
Agentic
8.30
Mercor APEX · 2026-10-18
Human preference
1351.76
Arena · Text style control (Elo) · 2026-10-18
Apache 2.0 · Arena