Back to articles

Exact model comparison

Mistral Small vs Ministral 8B vs Ministral 3B API pricing

These three exact Mistral routes are close enough to shortlist together, but their checked-in pricing and context fields point to different cost pressures. This page keeps to the database facts: token prices, context window, directional modality fields, source URL, and direct links into the live compare tool.

Updated June 8, 2026 Exact Mistral model IDs Built from checked-in data

At a glance

Three Mistral rows with different pricing shapes

Prices are shown in USD per 1M tokens from the current checked-in model rows. The source URL attached to all three rows is Mistral pricing, synced on 2026-05-28.

Mistral

Mistral Small

mistral/mistral/mistral-small-latest

Model details

Lowest input token rate in this set, with checked-in image and text input plus text output.

Input
$0.06
Output
$0.18
Context
131,072
Sample monthly
$11.40
Input: image + text
Output: text
Mistral

Ministral 8B

mistral/mistral/ministral-8b-2512

Model details

Largest checked-in context window in this set, with equal input and output token rates.

Input
$0.15
Output
$0.15
Context
262,144
Sample monthly
$19.50
Input: image + text
Output: text
Mistral

Ministral 3B

mistral/mistral/ministral-3-3b-2512

Model details

Lower output token rate than Mistral Small in this set, but the checked-in row has no reviewed directional modalities.

Input
$0.10
Output
$0.10
Context
131,072
Sample monthly
$13.00
Input: Not reviewed in checked-in row
Output: Not reviewed in checked-in row

Scenario

What the same request volume costs

Sample workload: 100,000 requests per month, 1,000 input tokens, and 300 output tokens per request.

Model Input / 1M Output / 1M Context Monthly cost
Mistral Small
mistral-small-latest
$0.06 $0.18 131,072 $11.40
Ministral 8B
ministral-8b-2512
$0.15 $0.15 262,144 $19.50
Ministral 3B
ministral-3-3b-2512
$0.10 $0.10 131,072 $13.00

How to choose

Pick by the field you need to protect

Choose Mistral Small for lower input spend

In this set, Mistral Small has the lowest checked-in input price. It also exposes image and text input in the checked-in modality fields, while still matching Ministral 3B's 131,072-token context window.

Choose Ministral 8B for context headroom

Ministral 8B is the context-window outlier here at 262,144 tokens. The tradeoff is a higher input token price than the other two rows in this comparison.

Inspect Ministral 3B when output price matters

Ministral 3B has a lower checked-in output rate than Mistral Small in this set. Its empty modality arrays should be treated as unreviewed database fields, not as a claim that the model lacks a modality.

Open the live comparison

Use the compare tool with all three exact model routes selected, then open model pages for source and metadata read-back.

Sources and caveats

Keep the comparison inside checked data

This article is a pricing and metadata read-back, not a quality, latency, benchmark, lifecycle, or availability ranking.

Mistral Small

mistral/mistral/mistral-small-latest

Ministral 8B

mistral/mistral/ministral-8b-2512

Ministral 3B

mistral/mistral/ministral-3-3b-2512

All three rows currently point to https://mistral.ai/pricing as their checked-in source URL. Confirm final provider billing before production use. This page does not add or change provider/model facts.

The sample cost excludes cache, batch, regional billing, tool-specific charges, overage rules, and any provider-side discounts. Empty modality arrays mean the checked-in row does not expose reviewed directional modalities.