These benchmarks should be taken very lightly. While, benchmarking is a good indication of one engine versus another, this doesn't show that, in reality, all of these template engines are very fast.
As a good rule, if your templating engine is your bottleneck, you shouldn't be using a templating engine.
Overall
Compile and Render Benchmark Scores (bigger is better). Run on a 2.93 GHz Intel Core i7 iMac with 12GB RAM.
Engine (npm name)
Average (mean)
Scores (bigger is better)
swig
106.23
111.02
104.62
106.65
103.77
107.21
107.61
105.95
101.54
107.66
jqtpl
39.44
40.28
40.04
38.79
40.03
38.51
40.43
38.60
38.55
39.70
whiskers
31.43
31.48
31.71
30.97
31.55
32.02
32.32
30.74
30.69
31.40
ejs
14.22
14.55
14.41
14.05
14.38
14.18
14.18
14.17
14.23
13.85
hamljs
5.55
5.47
5.54
5.54
5.57
5.65
5.65
5.51
5.51
5.49
jade
4.03
3.89
4.03
4.06
3.97
4.12
4.10
4.01
4.07
4.04
mu2
3.24
3.14
3.30
3.31
3.26
3.17
3.28
3.26
3.21
3.23
haml
1.94
1.99
1.97
1.98
1.95
1.94
1.93
1.88
1.89
1.89
templ8
1.66
1.63
1.67
1.66
1.66
1.70
1.70
1.68
1.65
1.61
coffeekup
0.80
0.77
0.80
0.81
0.80
0.80
0.81
0.82
0.78
0.78
Fastest
Versus
% Faster
Times Faster
Order of Magnitude Faster
swig complete
jqtpl
62.87%
2.69
0.43
coffeekup
99.25%
132.53
2.12
Compile Step
Compile-Only* Benchmark Scores (bigger is better). Run on a 2.93 GHz Intel Core i7 iMac with 12GB RAM.
Engine (npm name)
Average (mean)
Scores (bigger is better)
jqtpl
649.94
639.22
634.30
669.27
664.76
652.15
647.11
647.25
668.13
627.26
swig
605.57
618.64
579.60
623.79
630.16
599.51
599.52
595.87
597.47
605.59
whiskers
37.68
37.88
38.12
37.71
38.67
38.09
37.61
36.96
36.49
37.61
ejs
19.64
19.51
19.56
20.26
19.75
20.14
19.84
18.80
19.45
19.46
hamljs
6.11
6.38
6.48
6.49
5.99
6.16
5.98
5.40
6.05
6.09
jade
5.01
4.96
5.20
5.05
5.03
5.04
4.95
4.81
5.05
5.02
mu2
3.77
3.80
3.85
3.62
3.90
3.93
3.63
3.79
3.69
3.71
haml
2.04
2.03
2.12
2.09
2.02
2.07
2.07
2.03
1.97
1.92
templ8
1.90
1.90
1.89
1.97
1.93
1.90
1.93
1.79
1.92
1.88
coffeekup
0.87
0.87
0.89
0.87
0.86
0.87
0.88
0.87
0.86
0.88
Fastest
Versus
% Faster
Times Faster
Order of Magnitude Faster
jqtpl compile
swig
6.83%
1.07
0.03
coffeekup
99.87%
740.91
2.87
Render Step
Render-Only Benchmark Scores (bigger is better). Run on a 2.93 GHz Intel Core i7 iMac with 12GB RAM.
Engine (npm name)
Average (mean)
Scores (bigger is better)
whiskers
263.25
263.65
265.58
259.86
262.45
267.25
263.00
262.90
262.96
261.63
swig
120.26
121.07
117.53
117.49
118.75
121.85
119.18
120.53
122.23
123.73
haml
89.48
90.08
87.65
89.85
89.45
90.80
89.02
87.91
90.63
89.95
ejs
76.57
71.03
76.44
75.94
78.86
78.56
77.49
77.19
76.70
76.94
hamljs
59.05
58.16
57.79
58.58
60.12
59.53
59.53
58.97
59.10
59.65
jqtpl
41.77
42.09
40.27
40.67
40.27
42.81
42.09
42.25
42.51
42.95
mu2
37.32
36.82
36.64
36.60
37.85
37.51
38.06
37.19
37.54
37.64
jade
25.39
25.65
25.50
25.50
25.09
24.89
25.33
25.37
25.47
25.67
coffeekup
18.84
18.99
18.85
18.23
18.97
18.92
18.89
18.81
18.95
18.92
templ8
15.78
15.09
15.31
15.74
16.03
15.76
15.99
16.08
16.06
15.93
Fastest
Versus
% Faster
Times Faster
Order of Magnitude Faster
whiskers render
swig
54.32%
2.19
0.34
templ8
94%
16.68
1.22
To calculate these benchmarks on your own system, clone this repository and run node index.js from your command-line.