opendunes

Reflex

New

by MorphReleased 3 November 2026

This model takes in text, reads about 65 pages at once, and gives back short typed answers that can run to 58 pages. Its best use is putting a fixed list of questions to one long case file and getting a pick, a named thing or a score back for each. Questions asked again about the same file reuse the stored reading, so a batch of questions on one file costs less after the first. It does not handle pictures, sound or video, and it does not write essays, so jobs needing those call for a different kind of model.

TakesTextImages
GivesClassification
Create API keyCompare

What it costs

You pay for what each request uses. No subscription, no minimum.

≈ 0,014 DA per case3 000 DA buys about 210 000 cases

One case, the ordinary use

1 000 tokens sent× 13,98 DA per 1M0,014 DA
Total≈ 0,014 DA

Every rate

Meter↑ you send↓ comes back
Text tokens13,98 DA
Cached tokens read2,66 DA

per 1M tokens

Sending a 10-page file with a single question≈ 0,14 DA
Sending a full 65-page file with a single question≈ 0,91 DA
When you pay
After the answer, from your balance. Nothing is held.
Why it varies
The length of the text.
After each one
The exact charge shows on the request.
DepositDeposits from 50 DA, with Edahabia or CIB.

Where your request runs

One company runs this model. OpenDunes sends every request to it. DA per 1M tokens: ↑ what you send.

MorphUnited States…
↑ Sent, per 1M
13,98 DA
Latency
…
Provider↑ SentLatencyUptime
MorphUnited States13,98 DA……

You pay the price of the provider that served your request. Prices read 15:18 · speed over the last 30 minutes · uptime over 3 days.

Call it in code

One HTTP call, with your questions named as you like.

POST /alpha/decisions
Base URL
https://opendunes.com/v1
Model ID
morph/reflex
import os, requests
BASE = "https://opendunes.com"
KEY = os.environ["OPENDUNES_API_KEY"]
H = {"Authorization": f"Bearer {KEY}"}
resp = requests.post(f"{BASE}/alpha/decisions", headers=H, json={
    "model": "morph/reflex",
    "state": "The courier came two days late and the box was open.",
    "questions": {
        "is_complaint": {
            "type": "noul",
            "instructions": "Is it a complaint?",
            "criteria": {
                "true": "The customer is unhappy with the service.",
                "false": "The customer is asking a question or saying thanks.",
            },
        },
    },
})
print(resp.json())
pip install requestsFull reference for this route
Sign in to run it

Runs this request once on your balance.

Questions about Reflex

  • How much does Reflex cost in dinars?

    About 0,014 DA per case. ↑ 13,98 DA per 1M tokens.

  • Is Reflex free? Can I try it?

    No. You pay per use from your balance, with no subscription. Run it once from Call it in code, at the same price.

  • What is Reflex best for?

    This model takes in text, reads about 65 pages at once, and gives back short typed answers that can run to 58 pages. Its best use is putting a fixed list of questions to one long case file and getting a pick, a named thing or a score back for each. Questions asked again about the same file reuse the stored reading, so a batch of questions on one file costs less after the first. It does not handle pictures, sound or video, and it does not write essays, so jobs needing those call for a different kind of model.

  • How do I pay for Reflex without an international card?

    Deposit dinars with your Edahabia or CIB card. Every model, Reflex included, is paid from that balance. No foreign card, no conversion.

  • Does Reflex work in Arabic, French or Darja?

    Morph names no languages for Reflex. Arabic, French and Darja: not stated.

  • What can it answer?

    Yes or no, one choice from your list, or a score on your scale. Each answer comes with its probability.

  • Who makes Reflex, and when was it released?

    Morph. Released 3 November 2026.

Other Morph models, priced in dinars

See all Morph models
Morph V3 FastMorph · Jul 7, 2025

Morph V3 Fast does not write code from scratch, it applies an edit. You send an instruction, the original code and a short update marked in a fixed three-tag format, and it merges the update into the file and hands back the result at about 10 pages of code a second. That suits a coding assistant that proposes partial edits and needs them applied quickly. Morph V3 Large works at about 4 pages a second and is built for complex edits.

260 pages at onceFills your JSON
≈0,47 DAper message
View model
Morph V3 LargeMorph · Jul 7, 2025

Morph V3 Large applies code edits rather than writing code. You send an instruction, the original code and an edit marked in a fixed three-tag format, and it hands back the merged file at about 4 pages of code a second. It is the one built for complex edits, where the merge has to land precisely.

260 pages at onceFills your JSON
≈0,62 DAper message
View model