GPAI Ledger › C4AI Command A Plus (Cohere) › Capture 17 Sep 2026

Command A Plus — capture 20260917T061850Z

Filed under Cohere — C4AI Command A Plus, the source this project captured it from; the document itself is the filing of the model named above.

Verify: sha256sum 465ad10da2a73b55ca4291e34e665bf843f8f918fd8a0c4e8f0092a8d60f93ef.pdf must equal the hash above (the filename IS the expected hash); ots verify 465ad10da2a73b55ca4291e34e665bf843f8f918fd8a0c4e8f0092a8d60f93ef.pdf.20260917T061850Z.ots -f 465ad10da2a73b55ca4291e34e665bf843f8f918fd8a0c4e8f0092a8d60f93ef.pdf (opentimestamps.org) proves the bytes existed no later than the attestation time — an upper bound on the capture time; the fetch time above is the archive's own record (a freshly captured proof reports 'pending' here: the calendars anchor within hours, but this archive only upgrades the stored proof to its anchor on a later run, so expect a day or two). ots verify needs a local Bitcoin Core node (a pruned one is fine); without one, ots info on the proof prints the attesting block height and merkle path to check on any block explorer.

Extracted text

Machine-extracted text (layout may be lost; the authoritative content is the stored file above).

Public Summary
of Training Content
for Command A+
and family
Version of the Summary: v1.1
Last update: 10 September 2026
1. General information
1.1 Provider identification
Provider name and contact
details:
Cohere Germany GmbH
Contact: support@cohere.com, use subject line “AI Act Request”
Authorised representative
name and contact details:
n/a
1.2 Model identification
Versioned model name(s): Command A+ and North Small Translate
Public documentation is available at https:/ /docs.cohere.com/
Model dependencies: North Small T ranslate is based on the same base model as
Command A+
Date of placement of the
model on the Union market:
20 May 2026
PUBLIC SUMMARY OF  TRAINING CONTENT FOR COMMAND A + AND  FAMILY 1
1.3 Modalities, overall training data size and other characteristics
Modality Training data size Types of content
T ext Less than 1 billion tokens
1 billion to 10 trillion tokens
More than 10 trillion tokens
The model was trained on a diverse set of text
data sourced from publicly available content and
proprietary datasets accessed through partnerships
or developed by Cohere, including general web
content, technical documentation, code, and other
text across various fields such as education,
government, science, and technology.
Image Less than 1 million images
1 million to 1 billion images
More than 1 billion images
Command A+ was trained on a diverse corpus of
images sourced from publicly available content and
proprietary datasets, with a focus on enterprise-
relevant content such as tables, charts, pdfs, and
infographics. North Small T ranslate was not trained
with image content.1
Audio n/a n/a
Video n/a n/a
Other n/a n/a
1.3.1 Other characteristics
Latest date of data
acquisition/collection for
model training:
Up to June 2026
Description of the linguistic
characteristics of the overall
training data:
Multilingual, covering at least 48 languages, including all official
European Union languages.
Other relevant characteristics
of the overall training data:
The model is optimized and trained to excel at enterprise
relevant-tasks, such as coding, tool use, agentic capabilities,
multilingual tasks, document understanding, instruction-
following tasks, and long-context tasks.
Additional comments
(optional):
1Any references to image content as a modality in this Public Summary of T raining Content do not apply to
North Small T ranslate.
PUBLIC SUMMARY OF  TRAINING CONTENT FOR COMMAND A + AND  FAMILY 2
2. List of data sources
2.1 Publicly available datasets
Have you used publicly
available datasets to train the
model?
Yes
No
If yes, specify the modality(ies)
of the content covered by the
datasets concerned:
T ext
Image
Audio
Video
Other
List of large publicly available
datasets:
The training data for the model includes text from Common
Crawl (https:/ /commoncrawl.org/). Cohere employs various
techniques to curate data prior to using it in training to support
data quality and suitability for training, including deduplication,
filtering for toxic, harmful or otherwise unsuitable content, and
quality filtering.
General description of other
publicly available datasets not
listed above:
Other publicly available datasets include various sources of text,
and image content across fields such as education, government,
science, technology, that are relevant to coding, tool
use, multilingual tasks, document understanding, instruction-
following tasks, and long-context tasks. See the description
provided in Section 1.3 for details.
Additional comments
(optional):
2.2 Private non-publicly available datasets obtained from third parties
2.2.1 Datasets commercially licensed by rightsholders or their representatives
Have you concluded
transactional commercial
licensing agreement(s) with
rightsholder(s) or with their
representatives?
Yes
No
Other
If yes, specify the modality(ies)
of the content covered by the
datasets concerned:
T ext
Image
Audio
Video
Other
Additional comments
(optional):
Some of the third parties identified in Section 2.2.2 license their
datasets.
PUBLIC SUMMARY OF  TRAINING CONTENT FOR COMMAND A + AND  FAMILY 3
2.2.2 Private datasets obtained from other third parties
Have you obtained private
datasets from third
parties that are not
licensed as described in
Section 2.2.1, such as data
obtained from providers of
private databases, or data
intermediaries?
Yes
No
If yes, specify the modality(ies)
of the content covered by the
datasets concerned:
T ext
Image
Audio
Video
Other
If publicly known, list private
datasets obtained from other
third parties:
None of the datasets subject to this section 2.2.2 are publicly
known.
General description of non-
publicly known private
datasets obtained from third
parties:
Cohere partners with various third parties to source
multimodal and multilingual training data across fields such as
education, government, science, technology, that are relevant
to coding, tool use, multilingual tasks, document understanding,
instruction-following tasks, and long-context tasks. See the
description provided in Section 1.3 for details.
Additional comments
(optional):
2.3 Data crawled and scraped from online sources
Were crawlers used by the
provider or on behalf of?
Yes
No
If yes, specify crawler name(s)/
identifier(s):
Cohere makes information about its web crawlers available at
https:/ /docs.cohere.com/docs/cohere-web-crawlers.
Purposes of the crawler(s): Third parties may make use of crawlers in the process
of developing datasets identified in Section 2.2.2. Prior to
August 2025, Cohere collected certain web data using a crawler
bot that is no longer in use.
General description of crawler
behaviour:
It is Cohere’s policy to require that crawlers be designed to
respect robots.txt and other effective technological measures,
such as technological denial or restrictions of access like
paywalls or subscription models.
Period of data collection: Up to June 2026
PUBLIC SUMMARY OF  TRAINING CONTENT FOR COMMAND A + AND  FAMILY 4
Comprehensive description of
the type of content and online
sources crawled:
Multimodal training data across fields such as education,
government, science, and technology, that are relevant to
coding, tool use, multilingual tasks, document understanding,
instruction-following tasks, and long-context tasks. See the
description provided in Section 1.3 for details.
Type of modality covered: T ext
Image
Audio
Video
Other
Summary of the most relevant
domain names crawled:
The most relevant domains used to train the model include
publicly available content in a variety of languages, including
educational, government, science and technology, code, and
general-purpose content.
Additional comments
(optional):
2.4 User data
Was data from user
interactions with the AI model
(e.g. user input and prompts)
used to train the model?
Yes
No
Was data collected from
user interactions with the
provider’s other services or
products used to train the
model?
Yes
No
If yes, provide a general
description of the provider’s
services or products that were
used to collect the user data:
In most cases, Cohere’s customers use Cohere models in their
own environments or in third party environments, meaning
Cohere has no access to inputs submitted to its models. Where
permitted by a user via user controls and Cohere’s relevant
terms of service, de-identified data from the use of Cohere
models on Cohere-hosted environments (e.g. user inputs) may
be used in limited circumstances.
Type of modality covered: T ext
Image
Audio
Video
Other
Additional comments
(optional):
Additional information on user controls and our privacy
practices is available at:
https:/ /cohere.com/privacy
https:/ /cohere.com/enterprise-data-commitments
PUBLIC SUMMARY OF  TRAINING CONTENT FOR COMMAND A + AND  FAMILY 5
2.5 Synthetic data
Was synthetic AI-generated
data created by the provider
or on their behalf to train the
model?
Yes
No
If yes, modality of the synthetic
data:
T ext
Image
Audio
Video
Other
If yes, specify the general-
purpose AI model(s) used to
generate the synthetic data if
available on the market:
Cohere generated synthetic data using its own models, including
Command A, and other internal models.
Information about other AI
models, including provider’s
own AI model(s) not available
on the market, used to
generate synthetic data to
train the model to which this
Summary applies:
We may use other AI models to generate synthetic data
for various purposes, such as to generate examples, data
augmentation, and data evaluation.
Additional comments
(optional):
2.6 Other sources of data
Have data sources other than
those described in Sections
2.1 to 2.5 been used to train the
model?
Yes
No
If yes, provide a narrative
description of these data
sources and the data:
Cohere may work with vendors, annotators, and other third-
party experts to create or improve specialized datasets. For
example we work with specialized professionals to create or
improve training data that represents enterprise-relevant tasks
to optimize our model’s performance on such tasks.
Additional comments
(optional):
PUBLIC SUMMARY OF  TRAINING CONTENT FOR COMMAND A + AND  FAMILY 6
3. Data processing aspects
3.1 Respect of reservation of rights from text and data mining exception
or limitation
Are you a Signatory to
the Code of Practice for
general-purpose AI models
that includes commitments to
respect reservations of rights
from the TDM exception or
limitation?
Yes
No
Describe the measures
implemented before model
training to respect
reservations of rights from the
TDM exception or limitation
before and during data
collection, including the opt-
out protocols and solutions
honoured by the provider or,
as applicable, by third parties
from which datasets have
been obtained:
Cohere implements measures to respect applicable rights
reservations and opt-out signals relevant to the TDM exception.
This includes data deduplication and filtering techniques as well
as crawlers that are designed to respect robots.txt and not
bypass or circumvent effective technological measures, such as
technological denial or restrictions of access like paywalls or
subscription models.
Additional comments
(optional):
3.2 Removal of illegal content
General description of
measures taken:
Cohere employs various techniques to support data quality and
suitability for training, including deduplication, filtering for toxic,
harmful or otherwise unsuitable content, and quality filtering.
These measures are intended to reduce the inclusion of illegal
or unlawful content and improve the safety and reliability of
the training corpus. Cohere is also a member of the Internet
Watch Foundation, which supports efforts to identify, report and
remove child sexual abuse material online.
3.3 Other information (optional)
Other relevant information
about data processing
(optional):
PUBLIC SUMMARY OF  TRAINING CONTENT FOR COMMAND A + AND  FAMILY 7