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.

Flint (free) vs Extract: API price, context & features · api14

Recursive: Flint (free) vs ScaleDown: Extract

A side-by-side comparison of Recursive: Flint (free) (Free) and ScaleDown: Extract ($0.14 / 1M tokens) using current USD prices and capabilities from the api14 catalog. Both run through the same OpenAI-compatible API.

Specifications

	Recursive: Flint (free)	ScaleDown: Extract
Provider	Recursive	ScaleDown
Input	-	$0.14 / 1M tokens
Context window	262,144	1,000,000
Max output	235,929	900,000
Input	text	text, image, file
Output	text	text
Tool calling
Structured output
Reasoning control
Added	Oct 21, 2026	Oct 13, 2026

Key differences

ScaleDown: Extract reads 3.8× more context (1,000,000 tokens).
Only Recursive: Flint (free) supports tool calling.
Only Recursive: Flint (free) supports reasoning effort control.

In depth

Recursive: Flint (free) is a text model from Recursive, released on 2026-10-21. ScaleDown: Extract is a text model from ScaleDown, released on 2026-10-13. Both are served through api14 under the text category, and both take text in and return text out, so at the level of basic shape they are the same kind of endpoint. The difference is in what each one is built to do well and how much you can steer it.

ScaleDown: Extract has the larger context window of the two, at 1,000,000 tokens against 262,144 for Flint. It also accepts image and file inputs alongside text, while Flint accepts text only. Flint, on the other hand, exposes a much wider set of supported parameters: include_reasoning, logprobs, max_tokens, reasoning, response_format, seed, stop, temperature, tools, top_logprobs and top_p. ScaleDown: Extract supports only max_tokens. So Flint is the one with tool calling and reasoning controls, and it is the only one of the two with those capabilities. Neither model is flagged as cheaper on input or output, so treat them as about the same price and check the live table for the actual numbers.

If your task needs function or tool calls, or you want to turn reasoning on and read log probabilities, Flint is the pick, because ScaleDown: Extract does not offer those parameters at all. If your task is long document work, or you need to pass images or files in the same request, ScaleDown: Extract is the pick, because Flint cannot take those modalities and its context is roughly a quarter the size. For plain text generation where you just want a completion and control over sampling, either works, and Flint gives you more knobs (temperature, top_p, seed, stop) to shape the output.

Calling either model through api14 is the same operation. The API is OpenAI-compatible, so you point your existing client at the api14 base URL, send a chat completions request with your messages, and set the model field. To use Flint, set model to "recursive/flint-v1.0:free". To use ScaleDown: Extract, set model to "scaledown/extract". Nothing else in the request body has to change for a basic call. If you are moving from one to the other, swap the model id and, if you were relying on parameters the new model does not list, drop them, since ScaleDown: Extract only accepts max_tokens.

There is one case where the smaller, more restricted model is still the better pick: when the job is extraction over a large body of text, or over text plus images or files, and you do not need tools or reasoning output. ScaleDown: Extract covers that with its larger context and its extra input modalities, and you are not paying for controls you will not use.

Switch with one string

Both models use the same endpoint and request shape. Change the model value from "recursive/flint-v1.0:free" to "scaledown/extract" and prompts, streaming and tools keep working.

Recursive: Flint (free) details & playground →ScaleDown: Extract 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.chat.completions.create({
  "model": "scaledown/extract",
  "messages": [
    {
      "role": "user",
      "content": "Explain vector databases in two sentences."
    }
  ]
});
console.log(result.choices[0].message.content);

Recursive: Flint (free) vs ScaleDown: Extract

Specifications

Key differences

In depth

Switch with one string

More comparisons