Microsoft models
All 12 Microsoft models on OpenDunes, priced in Algerian Dinars.
12 models
Microsoft-Decision-1 reads the content you send and returns a probability for each of a fixed set of answer options instead of writing text. Use it where a pick-one call drives your app: sorting tickets, routing requests, checking an answer, or deciding whether to act, wait or flag the case for a person. It reads about 33 pages at once. It cannot chat, translate or summarize, so pair it with a text model for those jobs.
The low-latency text-to-speech from Microsoft AI for real-time voice, producing expressive speech in 23 languages for voice agents, assistants and call centres, where the full MAI-Voice-2.1 gives up speed for fidelity
Microsoft AI: MAI-Voice-2
Going away Dec 31, 2026#1 for realistic voice · MAI-Voice-2 turns text into speech and returns audio, with control over tone and delivery and a way to give different lines to different speakers. It covers 15 languages in 18 regional versions and offers four ready voices: US English, Mexican Spanish, French and German. Choose it for narration, tutorials and assistants where natural delivery counts more than speed.
Microsoft AI: MAI-Transcribe 1.5
Going away Dec 22, 2026Microsoft AI's speech recognition for captions, call records, subtitles and accessibility, reliable across 43 languages, many accents and noisy real-world audio, faster than version 1 on long files but without speaker labels
MAI-Voice-2.1 turns text into speech in 23 languages and returns audio from about 97 named voices. It is Microsoft's most expressive voice model, with detailed rhythm and emotion and a speaker who stays the same across a long recording. Use it for audiobooks, podcasts, lectures and brand audio where polish matters more than speed.
MAI-Transcribe 2 turns audio into a transcript in 60 languages, spots the language itself and follows speakers who switch languages mid-sentence. It adds speaker labels, word-by-word timing, keyword lists for specialist terms and a choice between word-for-word and cleaned-up text. On long recordings it runs quicker than 1.5, so pick it for meetings, calls and subtitle files, and keep 1.5 only for simple jobs without speaker labels.
Microsoft AI: MAI-Voice-2-Flash
Going away Dec 31, 2026#2 for realistic voice · MAI-Voice-2 Flash turns text into speech with little waiting, for voice assistants, call centres and other uses that must answer at once. It speaks 15 languages in 18 regional versions, lets you steer tone and delivery, and offers four ready voices. Microsoft rates it at twice the speed of MAI-Voice-2 with the same natural delivery.
#2 for photo editing · MAI-Image-2.6 Flash generates and edits images from a written description plus up to five reference pictures, with a shape you fix or leave to the model, and returns one image per request. It is the faster tier of its family, made for jobs that need many images in a short time at quality close to the full model.
#2 for pictures · MAI-Image-2.6 generates and edits images from a written description plus up to five reference pictures, and returns one image per request. Its strength is editing from several references: you blend a person, a product, a style and a setting while the composition and detail hold. You can fix the shape of the image or let the model choose it.
#5 for photo editing · Photorealistic and artistic images from a text prompt, with one optional reference image and eight aspect ratios or one chosen by the model; the standard MAI-Image-2.5, beside its Pro tier
MAI-Image-2.5 Pro takes a written description, with one optional reference picture, and returns one image per request in square, landscape or portrait shapes. Microsoft aims it at work where quality comes first: main banner images, detailed edits and precise text rendered inside the picture.
Phi-4 is a small text model from Microsoft Research, built to work through multi-step questions on modest computers with quick replies. You send up to about 16 pages of text and get text back. It works best on English input and follows instructions closely. Pick it for fast, light reasoning jobs in English, and a larger model when you need to read longer texts or work in other languages.