Mistral
Visit websiteAI models, agents, and development tools for hosted or private deployment.
Verified Nov 2, 2026
What is Mistral?
Mistral provides AI models and tools for building applications, customizing models, and using AI for everyday work. Its model range covers general tasks and coding, document processing with OCR, and audio and speech with Voxtral. Organizations can run models on their own infrastructure, in a private cloud, or through Mistral’s hosted services.
For developers and businesses, Forge supports model customization around internal data and requirements, while Studio provides tools to build, govern, and improve agents and applications. Studio also tracks API usage and spending across services such as completions, OCR, and fine-tuning. Hosted API billing varies by service, including token-based, page-based, and per-minute charges.
Vibe provides a user-facing assistant for chat, web search, coding, document uploads, image creation, and other tasks. It is available on the web, mobile, and through coding interfaces including a CLI. A free plan offers limited usage, while paid plans increase allowances and add collaboration or enterprise controls. Self-hosted model use is subject to licensing terms, including separate requirements for commercial deployments.
What can Mistral do?
General and specialist models
Choose models for general tasks, coding, document OCR, and audio or speech processing.
Model customization
Use Forge to tailor models to organizational data, use cases, and requirements.
Agent and application development
Build, govern, and improve AI agents and applications with Studio.
Flexible deployment
Run models on-premises, in your VPC, or on Mistral infrastructure with regional inference and dedicated compute options.
Vibe assistant
Use chat, search, coding, document uploads, and image generation through web, mobile, and coding interfaces.
API usage tracking
Monitor API calls, endpoints, user activity, and service costs in the Studio dashboard.
Who is Mistral for?
- Developers
- Build AI applications, access models through APIs, and use Vibe for coding tasks.
- Enterprise AI teams
- Customize models and deploy agents with private infrastructure and organizational controls.
- Individual professionals
- Use Vibe for research, document work, web searches, and everyday tasks.
How much is Mistral?
Official pricingFree
Free
Pro
$14.99/mo
Team
$24.99/mo
Enterprise
See website
Pro Education plan
$5.99/mo
Free plan: The Free plan includes limited Vibe chat, search, image creation and coding, plus access to Mistral Studio.
Frequently asked questions
Is there a free plan?
The Free plan includes limited Vibe access for chat, search, creation, and agentic coding in the terminal. It also includes access to Mistral Studio.
How is API usage billed?
Most models charge separately for input and output tokens. OCR is billed per 1,000 pages, speech models per minute, and tool APIs per call; batch processing and cached inputs can reduce applicable costs.
Can I self-host Mistral models?
Yes, models can run on your own infrastructure. Commercial deployments require a Mistral license, with separate terms for derivatives and production use.
Is there a student discount?
Verified students can get Mistral Pro for $5.99 per month, compared with the regular $14.99 monthly price.
What does Enterprise include?
Enterprise includes private deployments, custom models, agents and workflows, audit logs, SAML SSO, and white labeling. Custom SLAs and dedicated support are also available.
Mistral alternatives
Compare alternatives- 1
OpenAIAI models, ChatGPT, and an API platform for personal, business, and developer use. - 2
Stability AIGenerative AI models and tools for creating images, audio, video, and 3D assets. - 3
MiniMaxAI models, developer APIs, and apps for language, video, speech, and music. - 4
FLUXGenerate images, video, and audio through a browser playground or API. - 5
AutoGPTCoordinate AI specialists for research, content, sales, and recurring workflows.