A usage guide detailing per-token costs for Mistral Small, Medium, and Large models accessible via the hosted API.
Why it matters
Stork’s verdict on Mistral AI Pricing
Mistral AI Pricing reviewed by Stork AI · stork.ai/en/mistral-ai-pricing
overview
Mistral AI Pricing is an AI model pricing guide developed by Mistral AI that enables developers, businesses, and users of Mistral AI's hosted API to understand and calculate per-token costs for their large language models. It details the usage-based pricing for models such as Mistral Small, Mistral Medium, and Mistral Large, facilitating cost management for AI-powered applications.
features
Mistral AI Pricing provides a transparent framework for understanding the operational costs associated with deploying Mistral AI's advanced language models through its hosted API. Key features include:
use cases
Mistral AI Pricing is designed for various stakeholders who need to manage and forecast expenses related to AI model deployment and usage. Primary beneficiaries include:
how to use
To effectively utilize Mistral AI Pricing, users should consult the official documentation to understand the per-token costs and apply them to their anticipated usage scenarios.
https://docs.mistral.ai/usage/pricing/.pricing
Mistral AI operates on a usage-based pricing model, where costs are calculated per token for both input (prompt) and output (completion) generated by its hosted API models. Specific pricing tiers are available for different model capabilities.
Pricing Page
View Pricing→Similar Tools
Mistral AI's pricing strategy positions its models competitively within the generative AI market, offering a balance of performance and cost-effectiveness against established and emerging providers.
OpenAI offers a broad spectrum of models, from highly capable flagship models to more cost-effective options, and is a market leader in generative AI with extensive developer tools.
OpenAI's pricing varies significantly by model; for instance, GPT-5.4 Nano is priced at $0.20 input / $1.25 output per million tokens, making it competitive with Mistral's cheaper models, while its flagship GPT-5.5 Pro is considerably more expensive at $30 input / $180 output per million tokens. OpenAI also provides discounts for cached input tokens.
Anthropic focuses on developing safe and helpful AI, with its Claude models known for their long context windows and strong performance in reasoning and agentic tasks.
Anthropic's Claude models offer competitive per-token pricing; for example, Claude Sonnet 5 has an introductory price of $2 input / $10 output per million tokens. While generally higher than Mistral's Large 3 ($0.50 input / $1.50 output per million tokens), Claude models provide strong agentic capabilities and large context windows.
Google Cloud AI provides access to Google's multimodal Gemini models, known for their large context windows and deep integration with Google Cloud services.
Gemini's pricing, such as Gemini 3.1 Pro at $2 input / $12 output per million tokens for contexts up to 200K tokens, is competitive with other frontier models and offers a significant 2M-token context window. Its Flash models, like Gemini 3.1 Flash-Lite at $0.25 input / $1.50 output per million tokens, are more cost-effective for high-volume, lower-latency tasks.
Cohere specializes in enterprise-grade LLMs with a strong focus on retrieval-augmented generation (RAG) and long-document processing.
Cohere's Command models are positioned competitively for enterprise RAG applications; Command R+ is priced at $2.50 input / $10.00 output per million tokens. Its Command R7B model is one of the cheapest production APIs available at $0.0375 input / $0.15 output per million tokens, making it highly cost-effective for certain use cases.
AI Reputation Report
ChatGPT, Perplexity, Gemini, Claude & Grok answer buyer questions about Mistral AI Pricing every day. See whether they name Mistral AI Pricing — or send buyers to a rival.