Skip to content
System One

System One API. One endpoint.

Call System One models through the System One API for classification, scoring, and yes/no checks. Get structured choices, scores, and probabilities; your application controls every action.

Get an API keyPlayground samples

HTTP + JSON · API key authentication · Prepaid credits

https://system-one.dev/v1
cURL · POST /v1/systemone
curl 'https://system-one.dev/v1/systemone' \  -H "Authorization: Bearer $SYSTEM_ONE_API_KEY" \  -H 'Content-Type: application/json' \  --data '{    "state": { "message": "I was charged twice." },    "questions": {      "team": {        "type": "choice",        "instructions": "Who should handle this?",        "criteria": {          "billing": "Payments and refunds",          "support": "Technical issues"        }      }    }  }'

Start calling System One

  1. 1

    Create a key

    Create an API key in the console and save it in your server environment. The full key is shown once.

  2. 2

    Send a request

    Copy the cURL example, or use your HTTP client. No SDK installation is required.

  3. 3

    Review usage

    Check request status, credit charges, and your remaining balance in the console.

Set SYSTEM_ONE_API_KEY on your server, then send the request to the endpoint shown above. Successful evaluations use your account credits.

API documentation

Three primitives for structured decisions

Send the context as state and define the questions to evaluate. Combine Choice, Score, and Noul in one request; each question is evaluated independently against the same state.

SHARED STATE

“I was charged twice for this order. Please refund the duplicate payment.”
state.message
Illustrative example · not a live response

Choice: select an option

Define a set of options. Get the selected option, a probability for each option, and confidence.

01
billingBilling & refunds
0.92
supportTechnical support
0.06
salesSales
0.02
EXAMPLE ANSWERanswers.department.choice → "billing"
Choose between Choice, Score, and Noul

These values illustrate the response format. Run your own inputs in the playground for a real result.

Try this example live

Compare Clairic and decision models

This list shows every decision model you can call on system-one.dev, including Clairic. Each page says what inputs it can read, how many questions you can ask at once, and how many credits a call costs.

Jev

1×

TypeSafe Jev structured decisions.

  • 1 credit per question per 4KB
  • Up to 32 questions · 64KB per request
  • Inputs: text and JSON
  • Gateway ceilings: 64KB for text-only models; 13MB for image-capable models. The selected model’s request limit above may be lower and includes base64 images.
  • Pick one: up to 255 options; Score: 2–10 levels
  • Question IDs: up to 128 characters
  • Nullable state and optional nullable instructions. Nonblank question IDs.

Model page

Clef Flash

3×

Structured classification and routing for high-volume, cost-sensitive work.

  • 3 credits per question per 4KB
  • Up to 32 questions · 13MB per request
  • Inputs: text, JSON, and images (PNG, JPEG, WebP)
  • Images: up to 4 · 4MB each · 8MB total
  • Gateway ceilings: 64KB for text-only models; 13MB for image-capable models. The selected model’s request limit above may be lower and includes base64 images.
  • Pick one: up to 255 options; Score: 2–10 levels
  • Question IDs: up to 100 characters
  • State cannot be null. Every question needs nonempty string, object, or array instructions. IDs use A–Z, a–z, 0–9, _, ., and -. Score criteria cannot be null; Noul criteria must be omitted or an object with only true/false keys. Additional question fields are not supported.

Model page

Clef

6×

Structured decisions for demanding classification and scoring.

  • 6 credits per question per 4KB
  • Up to 32 questions · 13MB per request
  • Inputs: text, JSON, and images (PNG, JPEG, WebP)
  • Images: up to 4 · 4MB each · 8MB total
  • Gateway ceilings: 64KB for text-only models; 13MB for image-capable models. The selected model’s request limit above may be lower and includes base64 images.
  • Pick one: up to 255 options; Score: 2–10 levels
  • Question IDs: up to 100 characters
  • State cannot be null. Every question needs nonempty string, object, or array instructions. IDs use A–Z, a–z, 0–9, _, ., and -. Score criteria cannot be null; Noul criteria must be omitted or an object with only true/false keys. Additional question fields are not supported.

Model page

Clef Omni

4×

Text, JSON, and image decisions for classification, routing, and scoring.

  • 4 credits per question per 4KB
  • Up to 32 questions · 13MB per request
  • Inputs: text, JSON, and images (PNG, JPEG, WebP)
  • Images: up to 4 · 4MB each · 8MB total
  • Gateway ceilings: 64KB for text-only models; 13MB for image-capable models. The selected model’s request limit above may be lower and includes base64 images.
  • Pick one: up to 255 options; Score: 2–10 levels
  • Question IDs: up to 100 characters
  • State cannot be null. Every question needs nonempty string, object, or array instructions. IDs use A–Z, a–z, 0–9, _, ., and -. Score criteria cannot be null; Noul criteria must be omitted or an object with only true/false keys. Additional question fields are not supported.

Model page

Decider

1×

Text and JSON decisions for classification, routing, and scoring.

  • 1 credit per question per 4KB
  • Up to 32 questions · 64KB per request
  • Inputs: text and JSON
  • Gateway ceilings: 64KB for text-only models; 13MB for image-capable models. The selected model’s request limit above may be lower and includes base64 images.
  • Pick one: up to 255 options; Score: 2–10 levels
  • Question IDs: up to 128 characters
  • Nullable state and optional nullable instructions. Nonblank question IDs.

Model page

Clairic

2×

An extraction model that returns typed values from text, JSON and images.

  • Provided by: system-one.dev
  • 2 credits per 4KB of input
  • Field count does not change the price. Input includes state, field specs, and images. Output is not charged.

Model page

Prices are in credits. Token usage is shown for reference only; you are charged by model, number of questions and 4KB blocks.

Models and request limits

Evaluate up to 32 questions over one shared state. Text and JSON models default to 64KB requests. Image-enabled models accept inline PNG, JPEG, and WebP within their published limits, with a gateway ceiling of 13MB. Check the live catalog for each model’s capabilities and limits.

GET /v1/models is public and returns currently published models, prices and limits. Inference still requires a platform API key.

API documentation

Idempotency and retriesAPI errorsAPI versions and changes

Usage-based credit pricing

Requests are charged by model, question count, and input size. Estimate the credit cost in the Playground before running an evaluation.

questions × max(1, ⌈request bytes / model block size⌉) × model rate

API pricing

Use the TypeScript SDK

Define questions with choice(), score(), and booleanQuestion(). Connect with the same API key and get typed results in your application.

Explore the SDK

Connect SDK to hosted API

Make your first HTTP request

Set SYSTEM_ONE_API_KEY on your server, then send the request to the endpoint shown above. Successful evaluations use your account credits.

Get an API key