We use cookies. Cookies and browser storage keep you signed in and remember your preferences. With your permission we also measure how the site is used. You can any time. See our cookie policy.
morph/reflexRuns on your balance at the prices above. New accounts start with free usage.
Settings
Morph: Reflex is a decision model from Morph. It belongs to the decisions category, and its job is to answer questions about a state rather than to produce free-form prose. If you are building a system that needs a typed answer rather than a paragraph, this is the kind of model the page describes.
The model accepts text and images as input, and it returns decisions as output. Its context window is 65,536 tokens, with a maximum output of 58,982 tokens. The decision output is typed: it answers choice, noul and score questions about a state. Because the state is prefix-cached across requests, repeat questions on the same state are handled with the cached input. That combination of image and text input with a typed decision output is what shapes where the model fits.
Practical use follows from that shape. You can send a state, whether it is written out as text or supplied as an image, and ask a choice question to pick between options, a noul question, or a score question to get a graded judgement. Because the same state can be reused across requests through prefix caching, it suits workflows where one state is examined repeatedly with different questions, such as triage, routing, classification, or evaluation steps inside a larger pipeline. The image input means the state does not have to be transcribed first if it already exists as a screenshot or scan.
You call the model through api14's OpenAI-compatible API using the model id "morph/reflex". The request format follows the OpenAI chat completions convention, so existing OpenAI client libraries work with the base URL pointed at api14. This model lists no supported parameters, so you should not expect to pass sampling controls such as temperature or top_p; send the state and the question and read the decision from the response. The model id is the only identifier you need to select it.
Because the output is a decision rather than generated text, this model is not the right fit when you need a written explanation, a summary, or conversational replies. In those cases a general text generation model will serve you better. Reflex is aimed at the narrower job of answering typed questions about a state, and it is worth choosing when that is exactly what your application needs.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api14.com/api/v1",
apiKey: process.env.API14_API_KEY,
});
const result = await client.post("/decisions", { body: {
"model": "morph/reflex",
"state": "Customer: I was charged twice for my order and want my money back.",
"questions": {
"team": {
"type": "choice",
"instructions": "Which team should handle this message?",
"criteria": {
"billing": "Payments, refunds, invoices",
"support": "Product questions and bugs",
"sales": "Pricing and new purchases"
}
},
"urgent": {
"type": "noul",
"instructions": "Does this need a reply today?",
"criteria": {
"true": "Money, outage or legal risk",
"false": "Can wait"
}
}
}
} });
console.log(result.answers);
This community has no topics yet. Open the first one about Morph: Reflex.