MistralBase plan

Ministral 14B

Mistral AI's compact Ministral model via the Vercel AI Gateway: efficient text and vision at a 256k context window with flat, low pricing and function calling. Available at the base tier. Uses 1 base request per send before length multipliers. Supports image input. Does not support web search.

You always get the exact model you pick — we never silently route you to another.

About Ministral 14B

Ministral 14B is Mistral's compact, efficiency-minded model, and on just4o.chat it runs through the Vercel AI Gateway as mistral/ministral-14b. It is built to be fast and inexpensive while still handling both text and image input — a rare combination at its flat $0.20-per-million price for input and output alike. That symmetry makes its cost easy to reason about, and at the base tier it bills just 1 base request per send before length multipliers, so it is a natural choice for high-throughput tasks, lightweight assistants, and anywhere predictable low cost matters. A 256k-token context window is generous for a model this size, and function calling is supported for simple tool use. The caveats are what you would expect from a small model: no native web search, and it will not match the reasoning depth of Mistral Large 3, Medium 3.5, or Magistral on hard problems. For quick multimodal tasks, classification, and cost-efficient everyday work, it punches above its size.

Best for

  • High-volume, cost-sensitive tasks where flat $0.20 pricing is attractive
  • Lightweight multimodal work — quick reasoning over text plus images
  • Classification, extraction, and routing at scale
  • Simple tool-using assistants with function calling
  • Long-context tasks up to 256k tokens on a small-model budget

Specifications

ProviderMistral
Released2025-12
IntelligenceMedium
SpeedFast
Context window256,000 tokens
Knowledge cutoffNot officially published
Input price$0.20 / 1M tokens
Output price$0.20 / 1M tokens
Request cost1 base request
Plan tierBase
InputText, Image
OutputText
Features1 base request per send before length multipliers, Function calling supported, Image input supported, Served through the Vercel AI Gateway (mistral/ministral-14b), Mistral applies zero data retention to gateway traffic
Model IDministral-14b

Frequently asked questions

Yes. Despite its small size, it accepts both text and images, making it a budget-friendly multimodal option.

A flat $0.20 per million tokens for both input and output at provider reference pricing. On just4o.chat it is a base model that uses just 1 base request per send before length multipliers.

Through the Vercel AI Gateway's OpenAI-compatible endpoint under the mistral/ministral-14b route. Mistral applies zero data retention to gateway traffic.

256,000 tokens, generous for a compact model.

No native web search, and as a small model it does not match the reasoning depth of Mistral's larger models on complex problems. It shines on speed, cost, and lightweight multimodal tasks.