by MoondreamReleased 15 October 2026
Moondream 3 Preview takes a picture and a question in words and writes back an answer. It can describe a picture, caption it, read the text in a photo and work out what is going on, with an optional mode where it thinks before it answers. It is a small model, so it suits tagging or describing large numbers of pictures.
You pay for what each request uses. No subscription, no minimum.
One message, the ordinary use
Every rate
| Meter | ↑ you send | ↓ comes back |
|---|---|---|
| Text tokens | 99,85 DA | 832,09 DA |
per 1M tokens
One company runs this model. OpenDunes sends every request to it. DA per 1M tokens: ↑ what you send · ↓ what comes back.
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.
Real requests, real replies. Try any of them yourself — one click, pre-filled.
What's the total on this invoice, and when is it due?
invoice-scan.jpg · photoThe total is 42 300,00 DA, due 5 September 2026. Worth a double-check: the handwritten day could read as 8 instead of 3.
Try in Chat LabTwo lines to welcome a shop in Oran.
Welcome to our new shop in Oran — fresh bread every morning, and coffee the way the city likes it. Come in, the first cup is on us.
Try in Chat LabTry your own request — same model, pre-selected.
Open Chat LabSwap the base URL and the key. Your OpenAI code runs unchanged.
import os
from openai import OpenAI
client = OpenAI(
base_url="https://opendunes.com/v1",
api_key=os.environ["OPENDUNES_API_KEY"],
)
resp = client.chat.completions.create(
model="moondream/moondream3-preview",
messages=[{"role": "user", "content": "Two lines to welcome a shop in Oran."}],
)
print(resp.choices[0].message.content)Runs this request once on your balance.
Standard OpenAI fields, grouped by what they control. Anything not listed is accepted and ignored.
Shaping the reply
How long, how varied, where to stop.
max_tokenstemperaturetop_pReasoning
Let it think before answering — and see the thinking.
reasoninginclude_reasoningOther
Also accepted by this model.
max_completion_tokensAbout 0,52 DA per message. ↑ 99,85 DA · ↓ 832,09 DA per 1M tokens.
No. You pay per use from your balance, with no subscription. Try it in Chat at the same price.
Moondream 3 Preview takes a picture and a question in words and writes back an answer. It can describe a picture, caption it, read the text in a photo and work out what is going on, with an optional mode where it thinks before it answers. It is a small model, so it suits tagging or describing large numbers of pictures.
Deposit dinars with your Edahabia or CIB card. Every model, Moondream 3 Preview included, is paid from that balance. No foreign card, no conversion.
Moondream names no languages for Moondream 3 Preview. Arabic, French and Darja: not stated.
It reads text and images. Up to 32K tokens in one request, about 32 pages. It writes up to 3 000 words at once.
Yes, its weights are public. Open weights on Hugging Face
Moondream. Released 15 October 2026.
This model is designed to convert spoken language into written text. It accepts audio input and produces a complete transcription. It is a specialized version of a known speech processing model, optimized to improve the accuracy of word recognition. Because it is designed specifically for speech-to-text, it is not suitable for tasks requiring complex reasoning or image analysis.
Moondream 3.1 takes a picture and a question in words and writes back an answer: a description, a caption, the words read off a photo, or an explanation of what the picture shows. It has an optional thinking mode and keeps about 32 pages of text in mind alongside the picture. It is a small vision model for jobs like describing product photos or pulling the words off a screenshot.
From what its maker and two benchmark sites publish, and where it stands on OpenDunes. Read 7 Nov 2026.
Thinks first
Works through the problem before it answers.
Uses your tools No.
Can call functions your code provides.
Fills your JSON No.
Answers in the JSON shape you give it.
Searches the web No.
Can look things up online while it answers.
Cheaper on repeats No.
A prompt it saw a moment ago costs less the second time.
Open weights
The model's weights are public.
Point to every object matching a plain-language description in a photo
Reason about a chart or scene with each step tied to part of the image
Thinks first
Read the text off a receipt, sign or label from a single picture