Serverless GPU inference for ML models
Pay-per-millisecond API to run ML in production.
Pay-per-millisecond API to run ML in production.
Example inference request
DALL·E Mega
Python
Javascript
curl
1import requests
2
3response = requests.post(
4 url="https://api.pipeline.ai/v2/runs",
5 headers={"Authorization": "Bearer YOUR_API_TOKEN"},
6 json={
7 "pipeline_id": "pipeline_17ac3021b7674b10a6fbe3cb980ff57d",
8 "data": [
9 ["Mountain winds, and babbling springs, and moonlight seas"],
10 {"num_images": 9},
11 ],
12 },
13)
14
Cheaper than AWS or GCPReduced GPU usage with serverless
Up-to-date enterprise hardwareNVIDIA Ampere and Volta GPUs.
Save engineering timeWe handle the cloud infrastructure as you scale.
Unlimited requestsNo changes required as your product grows
Reduced cold startLow latency and reliable response times.
Rapid supportPersonal specialist help.
State-of-the-art AI models, one API call away.
Explore our list of AI models available as an API.State-of-the-art AI models, one API call away.
Explore our list of AI models available as an API.
Dall·E Mini
Generate images from a text prompt.
Dall·E Mega
Generate images from a text prompt.
GPT-J
High-quality text generations on-par with GPT-3
GPT-2 Large
High-quality text generations.
GPT-2 Medium
Medium-quality text generations.
GPT-Neo 2.7B
High-quality text generations.
GPT-Neo 1.3B
Medium-quality text generations.
GPT-Neo 125M
Fast text generations.
What will you build next?
Some examples of what you can do with Pipeline.
Text generation
Text generation
If you are looking for an AI model to generate text, our NLP models are the most suitable.
GPT-J
GPT-J 16-bit
GPT-2
GPT-2 L
GPT-2 M
GPT-NEO
GPT-Neo 2.7B
GPT-Neo 1.3B
GPT-Neo 125M
Image generation
Image generation
If you are looking for an AI model to generate images, our DALL·E models are the most suitable.
DALL·E
DALL·E Mega
DALL·E Mini