VERIFIED PRICECODING MODEL128KTEXT MODELCACHE 90% OFF
Codestral API Pricing
Codestral is Mistral's code-specialized model at $0.30/$0.90 per million tokens. $0.3/M input, $0.9/M output. The model card lists 128K context and FIM support for coding workloads.
Input - per 1M tokens
$0.30/M
Source Mistral active
Output - per 1M tokens
$0.90/M
Use for cost checks active
Cached input
$0.03/M
Cache 10% of input, per model -90%
Effective - agentic blend
$0.144/M
92/8 split - 82% cache
§ 01 / TERMINAL
Run the numbers.
Calculator pre-loaded with Codestral rates. Tweak spend, output mix, or cache hit rate to compare this model with nearby alternatives.
$ /mo
Workload split
Prompt cache hit rate
Tokens you can process
—
Words equivalent (English)
—
Effective rate
—
§ 02 / SCENARIOS
Real-world presets.
AGENT LOOP
Repo-wide bug fix
$0.013/task
LONG DOC
Reading 100-page contracts
$0.021/doc
RAG SUPPORT
Support agent ticket triage
$0.00121/ticket
ASSISTANT TURN
Research planning turn
$0.00274/turn
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 257
Words 38
Tokens (estimated) 51 tokens
Cost as input · uncached $0.00002 USD
Cost as output · uncached $0.00005 USD
Cost as cached input $0.000002 USD
| Model | Input /M | Output /M | Effective blended | Context | Best for |
|---|---|---|---|---|---|
| Codestral Current | $0.30 cache $0.03 | $0.90 | $0.144 agentic 92/8 | documented elsewhere — not on pricing page | Coding and software agents |
| Mistral Large 3 | $0.50 cache $0.05 | $1.50 | $0.241 +67% | documented elsewhere — not on pricing page | Open-weight multimodal work |
| Mistral Small 4 | $0.15 cache $0.015 | $0.60 | $0.0842 −42% | 256K | Low-cost Mistral workloads |
| Mistral Medium 3.5 | $1.50 cache $0.15 | $7.50 | $0.962 +6.7× | documented elsewhere — not on pricing page | Agentic and coding workloads |
| GPT-5.4 mini | $0.75 cache $0.075 | $4.50 | $0.541 +3.7× | 400K | Subagents and lightweight coding |
| Claude Sonnet 4.6 | $3.00 cache $0.30 | $15.00 | $2.05 +14× | 1M | Current Claude agents |
By comparison
Frequently asked.
Short answers for teams checking Codestral pricing, status, and migration choices.
Q · 01 Is Codestral still available? +
Codestral is listed as active/current in the pricing snapshot and was checked against the vendor page used for this entry.
Q · 02 How much does Codestral cost? +
Codestral is listed at $0.3/M input and $0.9/M output.
Q · 03 Is cached-input pricing included? +
Yes. Mistral publishes a per-model cached-input rate on its inference pricing page:
$0.03/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 What should teams compare it against? +
Use the shelf to compare nearby current models. Snapshot replacement:
current same-provider models.Q · 05 Why keep this page? +
These pages preserve live price checks, preview economics, and migration context in one consistent calculator format.