API

Billions of parameters on tap

Access language models that read billions of open web pages and teach themselves to understand individual words, as well as the overall meaning, sentiment and tone of sentences and entire documents.

Use the API to generate completions, distill text into semantically meaningful vectors, measure the semantic similarity between two passages, and more. Get state-of-the-art natural language processing without the need for expensive supercomputing infrastructure.

Powerful

Our models gain a rich understanding of language and the world by reading text across billions of web pages.

Customizable

Finetune our models with your data to create tailored models, purpose-built for your needs.

Multi-purpose

Access language generation, model representations, and conditional likelihoods, all under one API.

Easy Access

Effortlessly deploy language models with billions of parameters into your product.

Compose

Generate natural-sounding text from a prompt.

import cohere
co = cohere.CohereClient('YOUR_API_KEY')
predictions = co.generate(
model='baseline-shark',
prompt='Rephrase the following sentence:\n'
'1. When does the show start?\n'
'2. What time is the concert?\n'
'3. When is the beginning of the set?\n',
num_tokens=50)

Comprehend

Retrieve embeddings of text to understand its meaning.

import cohere
co = cohere.CohereClient('YOUR_API_KEY')
emb = co.embed(
model='baseline-seal',
texts=['This album is not that great.', 'Their old stuff was better.'])
if emotion_classifier.predict(emb) < 0.5:
print('They did not like it')

Compare

Measure semantic similarity between texts.

import cohere
co = cohere.CohereClient('YOUR_API_KEY')
similarities = co.similarity(
model='baseline-seal',
anchor='title: The Life and Times of Art Garfunkel',
targets=['genre: Music',
'genre: Math',
'genre: Science'])

Get access to the Cohere API

We are currently in private beta for select applications and technologists. Apply below to get on our waitlist.

Deploy language models with billions of parameters