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. Get started

Playground

The playground runs DecisionNode request shapes in your browser once you sign in, with no key. Build the state and questions, run them, read the answers, and copy the exact request as curl, Python or TypeScript.

on this page3 sections
  1. What it does
  2. Bring your own request
  3. Open an example by link

Your account, real answers

Running needs an account: signed out, Run asks you to create one and runs nothing, and your draft is kept through sign-up. Nothing runs until you press Run, and every run is a real decision by the API on your own workspace, billed like any API request, to its plan: the allowance first, then purchased credit. What you see (answers, evidence, headers, the model that answered) is exactly what the API returned. To call the API from your own code, send the same request with a key from the console.

What it does#

  • State

    What it does
    Type or paste the input, as Text or JSON. A JSON state is an object or an array; switch to Text to send a plain string
  • Images

    What it does
    Drop, paste or pick up to 16 images. They are sent in images with the state, and each tile shows its id and its number (Picture 1, Picture 2, ...), the two ways a question can refer to it; the id is the one that stays right when images are reordered
  • Questions

    What it does
    Build them with the visual builder (Choice, Score, Truth, Number, Point and Box rows), or switch to JSON and paste a whole questions object. JSON mode checks what you type the way the API does and names each problem by its path. A Point or Box row picks the image to look in (required with two or more images) and the units, pixels or normalized
  • Model

    What it does
    Pick DecisionNode or DecisionNode Flash; only model changes
  • Run

    What it does
    Press Run, or Cmd+Enter (Ctrl+Enter on Windows and Linux) from anywhere on the page
  • Answers

    What it does
    Each answer as a visual: distribution bars for a choice, the level scale for a score, a gauge for a truth, bars over the grid for a number, the point or the box drawn on its image with present and confidence
  • JSON

    What it does
    The exact response body and headers
  • Code

    What it does
    The same request as curl, Python, TypeScript or its JSON body, ready to copy into your code; the Code button beside Run opens it
  • Compare

    What it does
    On an example that has one, the same questions asked plainly (no state or a loose sentence, no criteria) beside the example's request, both run together: Compare with a plain question, under Run
  • Examples

    What it does
    Presets that load a complete request: two first lessons on stating your rule (is a hot dog a sandwich?), a lesson per question type, builds you could ship, and point and box builds, and knowledge builds that decide against sample documents the playground adds to your workspace (Knowledge)
  • Tour

    What it does
    A short guided tour of the four parts
The parts of the playground
PartWhat it does
StateType or paste the input, as Text or JSON. A JSON state is an object or an array; switch to Text to send a plain string
ImagesDrop, paste or pick up to 16 images. They are sent in images with the state, and each tile shows its id and its number (Picture 1, Picture 2, ...), the two ways a question can refer to it; the id is the one that stays right when images are reordered
QuestionsBuild them with the visual builder (Choice, Score, Truth, Number, Point and Box rows), or switch to JSON and paste a whole questions object. JSON mode checks what you type the way the API does and names each problem by its path. A Point or Box row picks the image to look in (required with two or more images) and the units, pixels or normalized
ModelPick DecisionNode or DecisionNode Flash; only model changes
RunPress Run, or Cmd+Enter (Ctrl+Enter on Windows and Linux) from anywhere on the page
AnswersEach answer as a visual: distribution bars for a choice, the level scale for a score, a gauge for a truth, bars over the grid for a number, the point or the box drawn on its image with present and confidence
JSONThe exact response body and headers
CodeThe same request as curl, Python, TypeScript or its JSON body, ready to copy into your code; the Code button beside Run opens it
CompareOn an example that has one, the same questions asked plainly (no state or a loose sentence, no criteria) beside the example's request, both run together: Compare with a plain question, under Run
ExamplesPresets that load a complete request: two first lessons on stating your rule (is a hot dog a sandwich?), a lesson per question type, builds you could ship, and point and box builds, and knowledge builds that decide against sample documents the playground adds to your workspace (Knowledge)
TourA short guided tour of the four parts

Bring your own request#

To try a request your code already sends, paste its state into the state editor (JSON mode for an object or an array) and its questions object into the questions editor in JSON mode, then run it with either model selected. The Code tab gives it back as curl, Python or TypeScript. Playground runs are your workspace's data, under its settings, like any API call (data settings).

Open an example by link#

/playground?example=<id> opens the playground with an example loaded, ready to run, so a link in your own docs or chat can start someone on the right request. The ids:

  • hot-dog-bun

    Kind
    Lesson
    What it loads
    Is a hot dog a sandwich? A split bun counts
  • hot-dog-slices

    Kind
    Lesson
    What it loads
    Is a hot dog a sandwich? Only two slices count
  • lesson-choice

    Kind
    Lesson
    What it loads
    Choice: name the commit
  • lesson-score

    Kind
    Lesson
    What it loads
    Score: how severe is this bug report
  • lesson-truth

    Kind
    Lesson
    What it loads
    Truth: was the customer charged twice
  • lesson-number

    Kind
    Lesson
    What it loads
    Number: count the cars in the frame
  • traffic-junction

    Kind
    Build
    What it loads
    Junction camera that times its own signals
  • people-count

    Kind
    Build
    What it loads
    Concourse camera that opens the next gate
  • storeroom-incident

    Kind
    Build
    What it loads
    Night camera that raises its own alarm
  • sorter-lane

    Kind
    Build
    What it loads
    Sorter that sends every package to its lane
  • support-triage

    Kind
    Build
    What it loads
    Support inbox that routes itself
  • content-moderation

    Kind
    Build
    What it loads
    Moderation that acts on every comment
  • agent-guardrails

    Kind
    Build
    What it loads
    Guardrail for agent tool calls
  • doc-classify

    Kind
    Build
    What it loads
    Classify a document from its pages
  • receipt-check

    Kind
    Build
    What it loads
    Receipt and invoice checker
  • listing-photo

    Kind
    Build
    What it loads
    Listing photo match
  • lead-scoring

    Kind
    Build
    What it loads
    Lead scoring for new signups
  • overdue-invoices

    Kind
    Build
    What it loads
    Statement reader that chases late invoices
  • age-gate

    Kind
    Build
    What it loads
    Age gate on a sign-up selfie
  • seller-check

    Kind
    Build
    What it loads
    Marketplace seller check
  • tutor-on-camera

    Kind
    Build
    What it loads
    Tutoring session: both on camera
  • member-check-in

    Kind
    Build
    What it loads
    Climbing hall check-in by face
  • public-figure

    Kind
    Build
    What it loads
    Name the public figure in a photo
  • invoice-match

    Kind
    Build
    What it loads
    Supplier invoice against its order
  • license-plate-camera

    Kind
    Build
    What it loads
    License plate camera
  • conveyor-pick

    Kind
    Point and box
    What it loads
    Robot arm that picks the right package
  • panel-defect

    Kind
    Point and box
    What it loads
    Quality check that marks the defect
  • fence-person

    Kind
    Point and box
    What it loads
    Perimeter camera that finds the climber
  • landing-marker

    Kind
    Point and box
    What it loads
    Drone that finds its landing pad
  • belt-pick-damage

    Kind
    Point and box
    What it loads
    Pick point and damage box on one belt frame
  • knowledge-claim

    Kind
    Knowledge
    What it loads
    Water damage claim against the claims policy
  • knowledge-rover

    Kind
    Knowledge
    What it loads
    Delivery rover against its operating manual
  • knowledge-refund

    Kind
    Knowledge
    What it loads
    Refund request against the refund policy
  • knowledge-moderation

    Kind
    Knowledge
    What it loads
    Reported post against the trust and safety policy
Example ids
IdKindWhat it loads
hot-dog-bunLessonIs a hot dog a sandwich? A split bun counts
hot-dog-slicesLessonIs a hot dog a sandwich? Only two slices count
lesson-choiceLessonChoice: name the commit
lesson-scoreLessonScore: how severe is this bug report
lesson-truthLessonTruth: was the customer charged twice
lesson-numberLessonNumber: count the cars in the frame
traffic-junctionBuildJunction camera that times its own signals
people-countBuildConcourse camera that opens the next gate
storeroom-incidentBuildNight camera that raises its own alarm
sorter-laneBuildSorter that sends every package to its lane
support-triageBuildSupport inbox that routes itself
content-moderationBuildModeration that acts on every comment
agent-guardrailsBuildGuardrail for agent tool calls
doc-classifyBuildClassify a document from its pages
receipt-checkBuildReceipt and invoice checker
listing-photoBuildListing photo match
lead-scoringBuildLead scoring for new signups
overdue-invoicesBuildStatement reader that chases late invoices
age-gateBuildAge gate on a sign-up selfie
seller-checkBuildMarketplace seller check
tutor-on-cameraBuildTutoring session: both on camera
member-check-inBuildClimbing hall check-in by face
public-figureBuildName the public figure in a photo
invoice-matchBuildSupplier invoice against its order
license-plate-cameraBuildLicense plate camera
conveyor-pickPoint and boxRobot arm that picks the right package
panel-defectPoint and boxQuality check that marks the defect
fence-personPoint and boxPerimeter camera that finds the climber
landing-markerPoint and boxDrone that finds its landing pad
belt-pick-damagePoint and boxPick point and damage box on one belt frame
knowledge-claimKnowledgeWater damage claim against the claims policy
knowledge-roverKnowledgeDelivery rover against its operating manual
knowledge-refundKnowledgeRefund request against the refund policy
knowledge-moderationKnowledgeReported post against the trust and safety policy

Each example's full request and answer, with the code that acts on it, is in Examples. The Examples gallery opens the same ids.

previousCommand linenextConsole, keys and members

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 it does
  2. Bring your own request
  3. Open an example by link