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.

Classify-1 vs Tev1 4B Experimental: API price, context & features · api14

ScaleDown: Classify-1 vs Together: Tev1 4B Experimental

A side-by-side comparison of ScaleDown: Classify-1 ($0.14 / 1M tokens) and Together: Tev1 4B Experimental ($0.15 / 1M tokens) using current USD prices and capabilities from the api14 catalog. Both run through the same OpenAI-compatible API.

Specifications

	ScaleDown: Classify-1	Together: Tev1 4B Experimental
Provider	ScaleDown	Together
Input	$0.14 / 1M tokens	$0.15 / 1M tokens
Context window	1,000,000	32,768
Max output	900,000	8
Input	text, image, file	text
Output	decisions	decisions
Tool calling
Structured output
Reasoning control
Added	Oct 13, 2026	Sep 30, 2026

Key differences

Both have about the same input price.
ScaleDown: Classify-1 reads 30.5× more context (1,000,000 tokens).
A request with 1,000 input and 500 output tokens costs $0.000140 with ScaleDown: Classify-1 and $0.000147 with Together: Tev1 4B Experimental.

In depth

ScaleDown: Classify-1 is a decision model from ScaleDown, and Together: Tev1 4B Experimental is a decision model from Together. Both sit in the decisions category, meaning their job is to return a decision rather than free-form prose, and both were released in late 2026, with Tev1 4B Experimental arriving on 30 September 2026 and Classify-1 on 13 October 2026. Neither model exposes any supported parameters, so you configure them purely through the prompt and the input you send.

The clearest split between them is context and input handling. Classify-1 accepts text, images and files, and it carries a context window of 1,000,000 tokens. Tev1 4B Experimental accepts text only and its context window is 32,768 tokens. On price, the two are about the same for input, so the choice is not really a cost question at the input side; output pricing is not something we can compare meaningfully here. Both return decisions as their only output modality, so neither is a general text generator.

For practical work, reach for Classify-1 when the material you need classified is large or mixed-format. Long contracts, scanned documents, screenshots, PDFs and image-heavy tickets all fit its input modalities and its very large context, so you can hand it a whole bundle in one call instead of chunking and stitching results yourself. Reach for Tev1 4B Experimental when the input is plain text and comfortably fits in 32,768 tokens, for example routing a support message, labelling a short record, or making a quick decision over a prompt you have already trimmed down. If your pipeline is text-only and your inputs are small, the smaller model is the simpler fit; the moment images, files or very long inputs enter the picture, Classify-1 is the one that can actually accept them.

Calling either model through api14 works the same way, because the gateway speaks the OpenAI-compatible chat completions format. You point your client at the api14 base URL, send your request as usual, and set the model field to the id you want. Switching between the two is a one-line change: use "scaledown/classify-1" for the ScaleDown model, or "togethercomputer/tev1-4b-experimental" for the Together model. Everything else in your request body, your headers and your parsing code stays as it is, so you can compare the two on your own data by swapping the string and re-running.

There are still cases where Tev1 4B Experimental is the better pick even though Classify-1 is the more capable on paper. If your inputs are short plain text and you want the leanest possible request, the text-only model avoids sending image or file payloads you do not need. It is also a reasonable default when you have already validated it on your task and simply want to keep that behaviour stable, since both models take no parameters and behave deterministically with respect to the prompt you give them.

Switch with one string

Both models use the same endpoint and request shape. Change the model value from "scaledown/classify-1" to "togethercomputer/tev1-4b-experimental" and prompts, streaming and tools keep working.

ScaleDown: Classify-1 details & playground →Together: Tev1 4B Experimental 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": "togethercomputer/tev1-4b-experimental",
  "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);

ScaleDown: Classify-1 vs Together: Tev1 4B Experimental

Specifications

Key differences

In depth

Switch with one string

More comparisons