Skip to content
DecisionNodeDecisionNdedocs
  • Guides
  • API reference
  • Examples
  • Playground

start here

  • QuickstartGet startedGet an API key, send one request with three questions, and branch your code on the typed answers. Plain HTTPS, nothing to install.
  • POST /v1/decideAPI referenceAnswer typed questions about a state, optional images and video frames. One request, one buffered JSON response, one answer per question.
  • QuestionsConceptsQuestions say what to decide. Each one has a type that fixes the shape of its answer: a choice from your options, a score on your scale, a…
  • ConfidenceConceptsProbabilities are calibrated per question type, so a threshold means what it says on the data we measure.
  • ImagesConceptsSend up to 16 images and text in the same request. The model reads printed and handwritten text, amounts, dates, objects and layout, and…
  • Batch jobsPatternsSend up to 10,000 requests in one file and collect the answers later, at a lower price than live calls.
  • Pricing and billingEvery workspace runs on a plan: Starter, Pro, Business or Dedicated.
↑↓ moveopen7 suggestions
Get API keyGet API key
DecisionNodeDecisionNde

Get started

  • Introduction
  • Quickstart
  • CLI
  • Playground
  • Console and keys
  • With coding agents
  • MCP server
  • Examples

Concepts

  • State
  • Questions
  • Choice
  • Score
  • Truth
  • Number
  • Points and boxes
  • Images
  • Video
  • Boosters
  • Collections
  • Confidence
  • Determinism
  • API keys

Knowledge

  • Overview
  • Documents
  • In a decision
  • Pricing and your data
  • When a rule does not appear

Models

  • DecisionNode
  • DecisionNode Flash
  • DecisionNode Edge
  • Limits
  • Versions
  • Dedicated capacity

Fine-tuning

  • Overview
  • Prepare your dataset
  • Upload and validation
  • Start a training run
  • Watch a run
  • The quality gate
  • Use your model
  • Limits and pricing

Patterns

  • Confidence-gated routing
  • Form router
  • Fan-out
  • Guardrails
  • Control loops
  • Batch jobs
  • Keys on devices
  • Bill your customers per device
  • Publishable keys

Integrations

  • Home Assistant
  • MQTT bridge
  • ROS 2

API reference

  • Overview
  • POST/v1/decide
  • POST/v1/systemone
  • GET/v1/models
  • POST/v1/uploads
  • /v1/keys
  • GET/v1/keys/{key_id}/usage
  • GET/v1/usage
  • POST/v1/credit
  • Errors
  • Safety check
  • Rate limits

Sessions API

  • Sessions overview
  • POSTOpen a session
  • WSStream frames
  • DELEnd a session

Collections API

  • Overview
  • POSTCreate a collection
  • GETList collections
  • GETRetrieve a collection
  • DELDelete a collection
  • POSTEnroll a face
  • GETList items
  • PATCHUpdate an item
  • DELDelete an item
  • POSTAdd a document
  • PUTReplace a document
  • GETRetrieve an item
  • GETList passages
  • DELDelete failed documents
  • POSTSearch a knowledge base

Batch API

  • The batch object
  • POSTCreate a batch
  • POSTAdd requests
  • POSTFinalize a batch
  • GETRetrieve a batch
  • GETGet batch results
  • POSTCancel a batch
  • GETList batches

Pricing and billing

  • Pricing and billing
  • Refer & earn

Policies

  • Responsible use
  • Data and privacy
  • Benchmarks
  • Pricing
  • Playground
Get API key
  • Guides
  • API reference
  • Examples
  • Playground

Get started

  • Introduction
  • Quickstart
  • CLI
  • Playground
  • Console and keys
  • With coding agents
  • MCP server
  • Examples

Concepts

  • State
  • Questions
  • Choice
  • Score
  • Truth
  • Number
  • Points and boxes
  • Images
  • Video
  • Boosters
  • Collections
  • Confidence
  • Determinism
  • API keys

Knowledge

  • Overview
  • Documents
  • In a decision
  • Pricing and your data
  • When a rule does not appear

Models

  • DecisionNode
  • DecisionNode Flash
  • DecisionNode Edge
  • Limits
  • Versions
  • Dedicated capacity

Fine-tuning

  • Overview
  • Prepare your dataset
  • Upload and validation
  • Start a training run
  • Watch a run
  • The quality gate
  • Use your model
  • Limits and pricing

Patterns

  • Confidence-gated routing
  • Form router
  • Fan-out
  • Guardrails
  • Control loops
  • Batch jobs
  • Keys on devices
  • Bill your customers per device
  • Publishable keys

Integrations

  • Home Assistant
  • MQTT bridge
  • ROS 2

API reference

  • Overview
  • POST/v1/decide
  • POST/v1/systemone
  • GET/v1/models
  • POST/v1/uploads
  • /v1/keys
  • GET/v1/keys/{key_id}/usage
  • GET/v1/usage
  • POST/v1/credit
  • Errors
  • Safety check
  • Rate limits

Sessions API

  • Sessions overview
  • POSTOpen a session
  • WSStream frames
  • DELEnd a session

Collections API

  • Overview
  • POSTCreate a collection
  • GETList collections
  • GETRetrieve a collection
  • DELDelete a collection
  • POSTEnroll a face
  • GETList items
  • PATCHUpdate an item
  • DELDelete an item
  • POSTAdd a document
  • PUTReplace a document
  • GETRetrieve an item
  • GETList passages
  • DELDelete failed documents
  • POSTSearch a knowledge base

Batch API

  • The batch object
  • POSTCreate a batch
  • POSTAdd requests
  • POSTFinalize a batch
  • GETRetrieve a batch
  • GETGet batch results
  • POSTCancel a batch
  • GETList batches

Pricing and billing

  • Pricing and billing
  • Refer & earn

Policies

  • Responsible use
  • Data and privacy
  1. docs
  2. /
  3. Policies

Data and privacy

What happens to the requests you send: how long they are kept, how to ask for zero retention, how your requests improve the models unless you opt out, and what sessions, batch jobs and the safety check hold.

on this page6 sections
  1. What is kept, and for how long
  2. Model improvement, and how to opt out
  3. Zero retention
  4. Personal data
  5. Sending less
  6. Contacts

The binding text is the policy

This page explains the Privacy Policy and the Data Processing Addendum for developers. Where they differ, the legal pages apply.

Bynn Intelligence's trust center lists the current attestations (SOC 2, ISO 27001:2022, ISO/IEC 42001:2023, GDPR, CCPA, PCI DSS SAQ A, NIST Cybersecurity Framework, HIPAA (BAA available), EU AI Act), monitored continuously in Vanta. Only the certifications and the SOC 2 report are attestations; the laws and frameworks listed are ones our programme is mapped to, not certifications. Protected health information may be sent only under a signed Business Associate Agreement.

Model improvement
Your requests and answers improve the models unless you opt out; never with zero retention, an Order Form or a Defence Contract Addendum
Retention
Requests and answers are deleted 30 days after each request
Zero retention
On request: requests and answers are not stored once the answer is returned
Sessions
The session's context is held for the life of the session only
Batch jobs
The request file is deleted when the results are written, or with the draft when it is not finalized within 24 hours; results are kept 7 days
Safety records
Never hold the text of a request
Your role
You are the controller of the data you send; we process it for you under the DPA

What is kept, and for how long#

  • Requests and answers (state, images, questions, answers)

    Kept
    30 days after each request
    Then
    Deleted. Not stored at all under zero retention
  • A session's context (its instructions, state, questions and recent frames)

    Kept
    For the life of the session, only to answer its later frames, also under zero retention
    Then
    Released when the session ends; its frames and replies are then kept as requests and answers
  • A batch's request file

    Kept
    Until the batch's results file is written, whatever the batch's end; a draft not finalized within 24 hours of its creation is cancelled and its request file deleted
    Then
    Deleted. An ended batch holds no request body
  • A batch's results

    Kept
    7 days after the batch ends
    Then
    Deleted; the batch's counters stay, and its results answer 410 results_expired
  • Requests kept for labelling, only while an Owner or Admin has Keep requests for labelling on: answered /v1/decide requests, their answers and their uploaded or inline images (not images sent by URL)

    Kept
    30 days after each request, encrypted, in your workspace's storage location, used only to build your own fine-tuning datasets
    Then
    Deleted; all at once when the setting is turned off, when zero retention is turned on, and with the workspace. Never kept under zero retention
  • Fine-tuning datasets and their images

    Kept
    Until 30 days after the dataset's last training run, unless you keep it
    Then
    Deleted, and whenever you delete the dataset or its model
  • Safety check records of flagged and refused requests

    Kept
    90 days after the request
    Then
    Deleted. They hold probabilities, the action taken and ids, never request text
  • Usage data (token counts, request ids, timings, the model that answered)

    Kept
    As long as billing and running the service need it
    Then
    Deleted or aggregated. It holds no request content
Data the API holds
DataKeptThen
Requests and answers (state, images, questions, answers)30 days after each requestDeleted. Not stored at all under zero retention
A session's context (its instructions, state, questions and recent frames)For the life of the session, only to answer its later frames, also under zero retentionReleased when the session ends; its frames and replies are then kept as requests and answers
A batch's request fileUntil the batch's results file is written, whatever the batch's end; a draft not finalized within 24 hours of its creation is cancelled and its request file deletedDeleted. An ended batch holds no request body
A batch's results7 days after the batch endsDeleted; the batch's counters stay, and its results answer 410 results_expired
Requests kept for labelling, only while an Owner or Admin has Keep requests for labelling on: answered /v1/decide requests, their answers and their uploaded or inline images (not images sent by URL)30 days after each request, encrypted, in your workspace's storage location, used only to build your own fine-tuning datasetsDeleted; all at once when the setting is turned off, when zero retention is turned on, and with the workspace. Never kept under zero retention
Fine-tuning datasets and their imagesUntil 30 days after the dataset's last training run, unless you keep itDeleted, and whenever you delete the dataset or its model
Safety check records of flagged and refused requests90 days after the requestDeleted. They hold probabilities, the action taken and ids, never request text
Usage data (token counts, request ids, timings, the model that answered)As long as billing and running the service need itDeleted or aggregated. It holds no request content

Image metadata such as EXIF is stripped when an image is read: only the pixels reach the model.

Model improvement, and how to opt out#

Unless you opt out, your requests and their answers may be used to train, evaluate and improve the models and the safety check, and to build datasets and models that we may license or sell. Identifiers and personal data are removed or reduced first where practicable, nothing is used to identify or contact anyone, personal data is never sold, and anything disclosed to a third party other than our own service providers is de-identified first. Your data is never used this way when zero retention is on, when you are on an Order Form or dedicated plan unless the Order Form allows it, or under a Defence Contract Addendum; safety records may always be used to improve the safety check. Opt out for the whole workspace in the console's settings, or by writing to privacy@bynn.com; it applies to requests sent after it takes effect. Usage data, which holds no request content, is used to run and improve the service. The binding text is section 8.4 of the Terms of Service and section 4 of the Privacy Policy.

Zero retention#

With zero retention, requests and answers are not stored once the answer is returned, and are never used to improve the models. It is offered on every plan: ask for it by writing to privacy@bynn.com with your workspace's name, and it applies to the keys of that workspace.

  • Sessions still hold their context for the life of the session, because each frame is answered on it. It is released when the session ends.
  • Batch jobs hold their request file until the results are written (a draft not finalized within 24 hours is cancelled and its file deleted), and the results for the days above. For data that must never be stored, send it through /v1/decide.
  • Usage data and safety records are kept as above; neither holds request text.

Personal data#

When your requests hold personal data, you are its controller and we process it on your behalf. The Data Processing Addendum is part of the terms and sets out the processor terms of Article 28 of the GDPR and similar laws: processing only on your instructions, security measures, breach notice, deletion, audits and transfers under the Standard Contractual Clauses.

  • Sub-processors fall into four categories: cloud computing and hosting, payment processing, email delivery, and error monitoring. The current list is in the console and on request, and we give 30 days' notice before adding one, so you can object.
  • Security: our information security management system is certified to ISO/IEC 27001:2022 and we hold a SOC 2 report from an independent auditor. The trust center has the certificate, the report and the other frameworks we follow; ask at security@bynn.com for anything it keeps under confidentiality.
  • Requests from the people in your data that reach us are forwarded to you, and we help you answer them.
  • When you leave, you can export the data we still hold for 30 days after the agreement ends; then it is deleted.

Sending less#

  • Send only what the decision needs. A ticket's text decides its route; the customer's name and address usually do not.
  • Replace direct identifiers with your own ids before they reach the state: customer_18342 instead of an email address.
  • Never put keys, passwords or card numbers in a state or an image.
  • For a consequential decision about a person, read Responsible use first: some need a qualified person's review.

Contacts#

Privacy questions, zero retention and data requests: privacy@bynn.com. Security issues: security@bynn.com.

previousResponsible use

DecisionNode is built and run by Bynn Intelligence, Inc.

  • Home
  • Playground
  • Examples
  • Console
  • Responsible use
  • Terms
  • Acceptable use
  • Privacy
  • Data processing
  • Defence addendum
  • Cookies
  • Affiliate program

on this page

  1. What is kept, and for how long
  2. Model improvement, and how to opt out
  3. Zero retention
  4. Personal data
  5. Sending less
  6. Contacts