by MistralReleased 6 October 2026
Mistral Large 4 takes text and pictures and answers in text, reading up to about a thousand pages and writing up to about 262 pages per reply. Its focus is reasoning, coding and automated work, with tool use, exact-form answers and a thinking-effort setting you can raise or lower per request. Pick it for whole-codebase or long-document jobs.
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 | 232,98 DA | 698,95 DA |
| Cached tokens read | 23,30 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.
| Provider | ↑ Sent | ↓ Back | Cache read | Latency | Throughput | Uptime |
|---|---|---|---|---|---|---|
| Mistral · Zdr | 232,98 DA | 698,95 DA | 23,30 DA | … | … | … |
| MistralEurope | 256,28 DA | 768,85 DA | 25,63 DA | … | … | … |
You pay the price of the provider that served your request. Prices read 15:20 · 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 LabTurn this customer email into JSON with intent, urgency and order_id.
customer-email.txtTry 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="mistralai/mistral-large-4-0",
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_pstopseedfrequency_penaltypresence_penaltylogprobstop_logprobsReasoning
Let it think before answering — and see the thinking.
reasoningreasoning_effortinclude_reasoningStructured output
Get JSON you can parse instead of prose.
response_formatstructured_outputsTools
Let the model call your functions.
toolstool_choiceAbout 0,58 DA per message. ↑ 232,98 DA · ↓ 698,95 DA per 1M tokens.
No. You pay per use from your balance, with no subscription. Try it in Chat at the same price.
Mistral Large 4 takes text and pictures and answers in text, reading up to about a thousand pages and writing up to about 262 pages per reply. Its focus is reasoning, coding and automated work, with tool use, exact-form answers and a thinking-effort setting you can raise or lower per request. Pick it for whole-codebase or long-document jobs.
Mistral Large 3 2512 costs 0,16 DA less for the same message.
Deposit dinars with your Edahabia or CIB card. Every model, Mistral Large 4 included, is paid from that balance. No foreign card, no conversion.
Mistral names no languages for Mistral Large 4. Arabic, French and Darja: not stated.
It reads text and images. Up to 1M tokens in one request, about 1 000 pages. It writes up to 197 000 words at once.
Yes, its weights are public. Open weights on Hugging Face
Mistral. Released 6 October 2026.
Mistral's flagship speech-to-text model that turns audio into transcripts with word timestamps and speaker labels, with context biasing for domain-specific vocabulary
The Voxtral Mini derivative that only transcribes, taking audio from meetings, voice notes and podcasts and returning the text through a standard transcription request
Mistral's compact speech-to-text model that turns audio into transcripts with word timestamps and speaker labels, and can be steered with context biasing to catch domain terms
Mistral AI's text embedding model from late 2023, whose 1024-dimension vectors capture how passages relate, for semantic search and RAG; Codestral Embed is the version for code
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
Can call functions your code provides.
Fills your JSON
Answers in the JSON shape you give it.
Searches the web No.
Can look things up online while it answers.
Cheaper on repeats
A prompt it saw a moment ago costs less the second time.
Open weights
The model's weights are public.
Reads image-and-text threads of up to 1M tokens and answers in structured JSON
Fills your JSON
Proves a flaw in open-source code is real, then writes the patch for it
Locates the part of an image that a question points to, a visual grounding task