Last verified
EDGE FLAGSHIP128K CONTEXTTEXT + VISIONOPEN WEIGHTS

Ministral 3 14B API Pricing

Ministral 3 14B is Mistral's current edge flagship, with live pricing verified from the official pricing page. Baseline rates are $0.2/M input and $0.2/M output; cached input is $0.02/M, 10% of input. Pulled from official Mistral sources and the project pricing snapshot.

Input - per 1M tokens
$0.20/M
Source Mistral flat
Output - per 1M tokens
$0.20/M
Context 128K flat
Cached input
$0.02/M
Cache 10% of input, per model -90%
Effective - agentic blend
$0.0642/M
92/8 split - 82% cache
§ 01 / TERMINAL

Run the numbers.

Live calculator pre-loaded with Ministral 3 14B rates. Tweak spend, output mix, or cache assumptions to compare it with current Mistral replacements.

$ /mo
Workload split
Prompt cache hit rate
Tokens you can process
—
Words equivalent (English)
—
Effective rate
—
Open full calculator (all models · share URL · CSV) →
§ 02 / SCENARIOS

Real-world presets.

CHATBOT

Support assistant

$0.0002/turn
2,000 in - 500 out~488,281 units/$100
CODE

Code completion

$0.00103/task
12,000 in - 2,000 out~97,200 units/$100
EDGE RAG

Compact retrieval

$0.00062/query
8,000 in - 1,000 out~161,498 units/$100
BULK

Batch classification

$0.00007/item
1,000 in - 100 out~1,381,215 units/$100
§ 03 / TOKENIZER

Paste text. See tokens. See cost.

What's a token →
Calibrated · measured on the vendor's tokenizer · 2026-06-10 Auto-counts as you type

Counts use a chars-per-token calibration measured on the vendor's own published tokenizer (mistralai/Mistral-Nemo-Instruct-2407, 2026-06-10). English prose is typically within a few percent; code and non-Latin scripts tokenize heavier. For billing-exact counts use the vendor's count-tokens API.

Characters 417
Words 68
Tokens (estimated) 82 tokens
Cost as input · uncached $0.00002 USD
Cost as output · uncached $0.00002 USD
Cost as cached input $0.000002 USD
§ 04 / SHELF

Up against the shelf.

All models →
Model Input /M Output /M Effective blended Context Best for
Ministral 3 14B Current $0.20 cache $0.02 $0.20 $0.0642 agentic 92/8 128K Edge inference and compact agents
Ministral 3 8B $0.15 cache $0.015 $0.15 $0.0482 −25% 128K Edge inference and compact agents
Ministral 3 3B $0.10 cache $0.01 $0.10 $0.0321 −2.0× 128K Edge inference and compact agents
Mistral Small 4 $0.15 cache $0.015 $0.60 $0.0842 +31% 256K Low-cost chat and support
Codestral $0.30 cache $0.03 $0.90 $0.144 +2.2× documented elsewhere — not on pricing page Coding and code agents
Mistral Large 3 $0.50 cache $0.05 $1.50 $0.241 +3.7× documented elsewhere — not on pricing page General Mistral workloads
DeepSeek V4.1 Flash $0.30 cache $0.006 $1.20 $0.15 +2.3× 1M Budget coding and bulk reasoning

Frequently asked.

Practical pricing questions, with archive status separated from current pricing.

Q · 01 What is Ministral 3 14B priced at? +
Ministral 3 14B is shown at $0.2/M input and $0.2/M output. This rate is the live baseline used in the calculator.
Q · 02 Is Ministral 3 14B still available? +
Ministral 3 14B is active in the current Mistral lineup, with pricing checked against the vendor pricing page on 2026-09-21.
Q · 03 Does this model have prompt-cache pricing? +
Yes. Mistral publishes a per-model cached-input rate on its inference pricing page: $0.02/M, 10% of the input rate. The calculator defaults to an 82% cache-hit rate, the assumption used across AI//COST; set it to 0% for workloads with no repeated prompt prefix.
Q · 04 How is the effective price calculated? +
AI//COST uses the same agentic blend everywhere: 92% input, 8% output, with 82% of input tokens read at the published cached rate. For Ministral 3 14B, that is $0.0642/M.
Q · 05 Why keep an archive page? +
Archive pages are useful for invoice checks, migration planning, and comparing how model pricing compressed over time. They are clearly labeled so retired API rows do not look like current recommendations.
Q · 06 Are there regional surcharges? +
The Mistral public pricing page is denominated in USD and EUR. This page stores the USD baseline; currency conversion and localized calculator variants are handled elsewhere on AI//COST.
Y
Reviewed by Yaroslav Vikhariev Founder · AI//COST · Prices checked by hand against docs.mistral.ai · Last verified Sep 21, 2026
Methodology Report a correction More by Y.V.