Ministral 8B
Mistral AI
DISCONTINUED. Mistral's edge-focused 8B model (October 2024), part of les Ministraux, built for on-device and low-latency inference with a 128K context window. It offers strong reasoning and instruction following for its size, suited to local assistants, function calling, and privacy-sensitive deployments. Mistral deprecated the hosted model on December 2, 2025 and retired it on December 31, 2025; the weights remain on Hugging Face under the Mistral Research License.
Strengths
- Edge-focused 8B model with a 128K context window
- Strong reasoning and instruction following for its size
- Good fit for on-device and low-latency inference
- Function calling for local agentic use
Caveats
- DISCONTINUED as a hosted API - Mistral retired it on December 31, 2025
- Weights remain under the Mistral Research License; Mistral names Ministral 3 8B as the replacement
- Research license for weights (commercial license required for production)
- 8B capacity trails larger models on complex tasks
- Text-only (no vision or audio input)
- Vendor
- Mistral AI
- Context Window
- 131,072 tokens
- License
- Mistral Research License
- Release Date
- 2024-10-16
- Modalities
- text
Capabilities
Resources
Pricing
Weights available under the Mistral Research License (commercial license available; the vendor priced the hosted endpoint at "$0.1 / M tokens (input and output)"). No longer served on the Mistral API: ministral-8b-2410 was deprecated 12/2/2025 and retired 12/31/2025 (replacement: Ministral 3 8B).