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.

New here? See how it works A 1-minute tour of the playground, models and your API key.

Microsoft-Decision-1 vs Drex v2.0: API price, context & features · api14

Microsoft: Microsoft-Decision-1 vs Nace.AI: Drex v2.0

A side-by-side comparison of Microsoft: Microsoft-Decision-1 ($0.15 / 1M tokens) and Nace.AI: Drex v2.0 ($0.21 / 1M tokens) using current USD prices and capabilities from the api14 catalog. Both run through the same OpenAI-compatible API.

Specifications

	Microsoft: Microsoft-Decision-1	Nace.AI: Drex v2.0
Provider	Microsoft	Nace.AI
Input	$0.15 / 1M tokens	$0.21 / 1M tokens
Context window	32,768	131,072
Max output	29,491	117,964
Input	text	text
Output	decisions	decisions
Tool calling
Structured output
Reasoning control
Added	Oct 9, 2026	Oct 12, 2026

Key differences

Microsoft: Microsoft-Decision-1 is 30% cheaper on input tokens.
Nace.AI: Drex v2.0 reads 4.0× more context (131,072 tokens).
A request with 1,000 input and 500 output tokens costs $0.000147 with Microsoft: Microsoft-Decision-1 and $0.000210 with Nace.AI: Drex v2.0.

In depth

Microsoft: Microsoft-Decision-1 is a decision model from Microsoft, listed under the decisions category. Nace.AI: Drex v2.0 is a decision model from Nace.AI, listed in the same category. Both take text as input and both return decisions as output, so they occupy the same slot in a pipeline: you send a description of a situation and you get a decision back rather than a chat reply or a block of generated prose. Microsoft-Decision-1 was released by its provider on 2026-10-09, and Drex v2.0 followed on 2026-10-12, so neither has a long track record behind it yet.

The clearest split between them is context. Drex v2.0 accepts up to 131072 tokens, four times the 32768 tokens that Microsoft-Decision-1 accepts. If your decision depends on a long document, a large batch of records, or a long history of prior steps, Drex v2.0 can hold all of that in a single call while Microsoft-Decision-1 would force you to trim or split the input. On modalities they are identical: text in, decisions out, and nothing else. The comparison data lists no tool calling, no structured output mode and no separate reasoning mode that belongs to only one of them, so you should not expect either model to expose a capability the other lacks. On price, Microsoft-Decision-1 is the cheaper of the two on input. The data does not establish a difference on output pricing, so treat output cost as roughly comparable and check the live price table for the numbers that apply to your traffic.

For day to day work, the choice usually comes down to how much text the decision needs. Short, self contained decisions, such as classifying a support ticket, picking a routing target, or choosing between a small set of options from a brief prompt, fit comfortably inside 32768 tokens and are the natural fit for Microsoft-Decision-1, which is also the cheaper option on input. Reach for Drex v2.0 when the decision has to be made over a long contract, a full log, a large candidate set, or a conversation that has been running for a while, because the larger window lets you pass the whole thing without preprocessing. If your inputs are short but very numerous, the input price advantage of Microsoft-Decision-1 matters more; if your inputs are long but few, the extra context of Drex v2.0 is worth more than the price difference.

Calling either model through api14 works the same way, because the gateway exposes an OpenAI-compatible API. You point your existing OpenAI client at the api14 base URL, send your key, and set the model field. The only change between the two is that string: use "microsoft/microsoft-decision-1" for the Microsoft model and "nace-ai/drex-v2.0" for the Nace.AI model. Your request body, your text input and your response handling stay as they are, so you can keep one code path and switch models with a configuration value, which makes it easy to route short inputs to one and long inputs to the other, or to compare their decisions on the same input before committing.

There is one case where the more expensive input side is still the better pick: when the decision genuinely needs the whole document in view. If trimming the input would drop evidence the decision depends on, Drex v2.0 is the right call even though Microsoft-Decision-1 is cheaper on input, because a cheaper decision made on incomplete text is not a saving. Conversely, when the input is short and the volume is high, Microsoft-Decision-1 remains the sensible default.

Switch with one string

Both models use the same endpoint and request shape. Change the model value from "microsoft/microsoft-decision-1" to "nace-ai/drex-v2.0" and prompts, streaming and tools keep working.

Microsoft: Microsoft-Decision-1 details & playground →Nace.AI: Drex v2.0 details & playground →

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": "nace-ai/drex-v2.0",
  "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);

Microsoft: Microsoft-Decision-1 vs Nace.AI: Drex v2.0

Specifications

Key differences

In depth

Switch with one string

More comparisons