OPEN MODEL
Mistral AI's family of language models, from small efficient models to large multimodal ones. Many are released as open weights for self-hosting, mostly under the Apache licence, and Mistral also offers a paid API, the Le Chat assistant and enterprise tools for custom deployments.
- PRICE
- Freemium
- SET-UP
- Advanced
- DATA
- Can run onshore
- FOUNDED
- 2023
- HQ
- Paris, France
- WEBSITE
- mistral.ai
- REVIEWED
- October 2026
SPEC SHEET
001 / WHAT IT DOES
- 01Open-weight models under the Apache licence for many releases
- 02Le Chat assistant with web search, document upload and agents
- 03API with per-token billing and fine-tuning options
- 04Enterprise deployment on private cloud or on-premises infrastructure
- 05Specialist models for coding and document OCR
BEST FOR
- Organisations wanting a European alternative to US model providers
- Teams self-hosting efficient models on limited hardware
- Enterprises needing on-premises deployment of a capable model
TYPICAL USES
002 / THE HONEST TAKE
STRENGTHS
- +Apache-licensed open weights for many models, including large ones
- +Choice of self-hosting, API or managed cloud access
- +Efficient smaller models suit limited hardware
- +Enterprise options for private and on-premises deployment
WATCH OUTS
- –Smaller third-party ecosystem than the largest US providers
- –Not every model is open, and some open ones use custom licences
- –Large models still need serious GPU capacity to self-host
003 / FOR AUSTRALIAN BUSINESSES
Not by default, but it can be kept in Australia by self-hosting or using it through an Australian cloud region.
Mistral's own API and Le Chat are hosted primarily in the European Union, and Mistral does not publish an Australian hosting region. Business tiers offer terms that exclude content from model training. For onshore processing, run open-weight models in an Australian cloud region, use the Mistral models AWS Bedrock offers in Sydney, or arrange a private deployment through Mistral's enterprise team.
PRIVACY ACT & PRACTICAL NOTES
Mistral is a French company, so its own hosted services operate under European privacy rules, but that does not automatically satisfy Australian requirements. For onshore processing, run the open-weight models in an Australian cloud region, or use the Mistral models AWS Bedrock offers in its Sydney region. Self-hosting keeps prompts under your organisation's control, which simplifies Privacy Act 1988 compliance for customer and staff data.
004 / WHAT IT COSTS
Freemium
Open-weight models are free to download and run; your costs are compute. Mistral's API is billed per token, the Le Chat assistant has free and paid tiers, and enterprise deployments are priced by agreement. See mistral.ai/pricing.
Check current prices on mistral.ai ↗︎005 / CONNECTS WITH
Want it wired into your CRM, inbox or Xero? That's what we build.
006 / ALTERNATIVES
Also worth a look
- LlamaOnshore possibleFree→
- QwenOnshore possibleFreemium→
- DeepSeekOnshore possibleFreemium→
- ClaudeOnshore possibleFreemium→
- Hugging FaceOnshore possibleFreemium→
- Google AI StudioOnshore possibleFreemium→
- OpenRouterOffshoreFreemium→
- OllamaOn deviceFreemium→
- LangChainAU optionFreemium→
- CohereOnshore possibleFreemium→
007 / QUESTIONS
Anything else is quicker to answer on a call. Thirty minutes, free, with the people who would build it.
Yes. Mistral's Le Chat assistant can be used in Australia on free and paid tiers, through the web and mobile apps. Consumer tiers may use conversations for training unless you opt out in settings, so businesses should use a business tier or the API with appropriate terms.
We set up and connect tools like this for Australian businesses, with the privacy settings done right from day one.
Book a free call→