Mistral Small 4
Mistral Small 4 is the latest iteration of Mistral AI's Small model line, delivering improved instruction following and stronger benchmark performance within a lightweight parameter footprint. Mistral positions Small 4 as their go-to model when cost-efficiency is paramount but output quality must remain competitive with models several times its size.
Compared to its predecessors, Small 4 shows tighter adherence to complex system prompts, better handling of multi-step tasks, and reduced hallucination on factual queries. It targets production workloads that require high call volumes — classification, extraction, summarization, and customer-facing chat — where per-token cost matters as much as quality.
Key Features
Enhanced instruction following over previous Mistral Small versions
Strong performance-per-dollar ratio for high-volume production workloads
Low latency suitable for real-time conversational applications
Function-calling and structured JSON output support
Multilingual capability across major European and global languages
Available via Mistral API and La Plateforme with straightforward pricing
Ideal Use Cases
High-volume classification and entity extraction pipelines
Customer support automation requiring fast, accurate responses
Summarization of business documents at scale
Chatbot backends where cost per conversation is a key metric
Rapid prototyping before scaling up to Mistral's larger model tiers
Example Prompts for Mistral Small 4
Technical Specifications
| Provider | Mistral |
| Category | Text |
| Modality | Text -> Text |
Frequently Asked Questions
Try Mistral Small 4 now
Start using Mistral Small 4 instantly — 100 free credits, no credit card required. Access 750+ AI models through one platform.