Replicate
Replicate.com is a mature platform for deploying and sharing AI models.
Replicate Serverless GPU Prices
CPU | 0 | 8 | 4 | 0.0001/s 0.36/h 0 | |
| 16 | 8 | 4 | 0.000225/s 0.81/h 0 | ||
| 48 | 16 | 4 | 0.000575/s 2.07/h -12% | ||
| 48 | 72 | 10 | 0.000725/s 2.61/h +12% | ||
| 48 | 65 | 10 | 0.000975/s 3.51/h 0 | ||
| 40 | 72 | 10 | 0.00115/s 4.14/h -10% | ||
| 80 | 144 | 10 | 0.0014/s 5.04/h +10% | ||
| 384 | 680 | 48 | 0.0058/s 20.88/h |
Recent benchmarks
14.04s 12:02:08 AM
105.98s 12:03:12 AM
16.03s 12:03:20 AM
13.13s 12:01:36 AM
11.5s 12:02:23 AM
16.1s 12:01:50 AM
13.63s 12:02:15 AM
11.17s 12:02:32 AM
12.29s 12:05:47 AM
12.95s 12:05:13 AM
29.15s 12:05:46 AM
68.23s 12:05:27 AM
15.13s 12:01:53 AM
14.73s 12:03:57 AM
94.01s 12:05:21 AM
13.8s 12:01:26 AM
12.97s 12:01:21 AM
14.81s 12:01:22 AM
18.4s 12:01:34 AM
28.58s 12:02:05 AM
21.91s 12:01:29 AM
16.72s 12:01:37 AM
19.43s 12:02:09 AM
27.27s 12:02:40 AM
144.54s 12:03:12 AM
11.56s 12:02:05 AM
14.29s 12:02:06 AM
17.63s 12:01:53 AM
14.34s 12:01:27 AM
14.8s 12:03:22 AM
11.19s 12:02:37 AM
14.76s 12:01:55 AM
hello_torch by replicate on t4
Median92.89s
Average125.92s
Failures0
92.93s 12:02:08 AM
142.32s 12:03:12 AM
164.9s 12:03:20 AM
81.48s 12:01:36 AM
142.46s 12:02:23 AM
98.45s 12:01:50 AM
78.77s 12:02:15 AM
96.77s 12:02:32 AM
296.68s 12:05:47 AM
269.82s 12:05:13 AM
58.57s 12:05:46 AM
294.94s 12:05:27 AM
80.73s 12:01:53 AM
201.34s 12:03:57 AM
275.8s 12:05:21 AM
85.29s 12:01:26 AM
79.32s 12:01:21 AM
79.23s 12:01:22 AM
82.83s 12:01:34 AM
92.85s 12:02:05 AM
88.39s 12:01:29 AM
81.64s 12:01:37 AM
84.34s 12:02:09 AM
147.23s 12:02:40 AM
118.57s 12:03:12 AM
82.26s 12:02:05 AM
87.08s 12:02:06 AM
82.01s 12:01:53 AM
84.28s 12:01:27 AM
144.06s 12:03:22 AM
135.36s 12:02:37 AM
98.84s 12:01:55 AM
About Replicate
There are large amounts of models made by the community as a showcase and for demonstration on the platform. You can publish your own model on Replicate for free, but you pay per second used by server.
With public models you only pay for the time the server is running, and with private models you also pay for the start up time.
You develop functions using the COG SDK, which makes it quite easy to containerize and deploy your AI models.
Replicate has stable support for webhooks, logging, and offer some observability.
The start up time for functions on Replicate is quite long (as of March 2024), so if fast responses are needed, it's better to use another vendor or configure Replicate to always keep a replica online.