A nimble French model that fits on a single modern GPU but still codes and writes decently.
Price per 1 million tokens
per 1M tokens you send
per 1M tokens you receive
Input tokens represent the text you send to the model for processing. Output tokens represent the model's generated response. Pricing is set by Mistral AI and reflects current API rates as of October 2025.
per month
per month
per month
Mistral 3.1 Small by Mistral AI is a capable general‑purpose AI model for chat, writing, and task automation.
Mistral AI
Text
Text-only processing
128000 tokens
Maximum tokens per request
$0.10
per 1M input tokens
$0.30
per 1M output tokens
Mistral 3.1 Small is a large language model from Mistral AI designed for real‑world applications where speed, quality and cost all matter. It’s priced at $0.10 per million input tokens and $0.30 per million output tokens, so teams can estimate usage‑based spend with simple math.
The model supports a context window of about 128,000 tokens, which is enough for long chats, multi‑document prompts, or passing rich system instructions without constant truncation.
For general product teams, it’s an easy default: fast enough for UI interactions, capable enough for summarization and drafting, and affordable for scale.
In stack choices, many teams compare Mistral 3.1 Small against ChatGPT 4.1 mini. The trade‑off usually comes down to latency tolerance, budget per request, and whether you need image understanding or deeper chain‑of‑thought style reasoning. If you’re cost‑sensitive, keep prompts concise, cache system instructions, and stream outputs so users perceive faster response. If quality is the priority, add brief exemplars and explicit success criteria in the prompt; small guidance often yields outsized gains.
For production, pair the model with guardrails (content filters, schema validators) and log prompts/responses for offline evaluation. Finally, create simple comparison tests—five to ten representative tasks from your app—to verify this model’s answers, latency and cost against your alternatives before you commit. To control spend, consider tiering workloads: route routine queries to a cheaper sibling and reserve this model for complex or customer‑visible moments. Add retries with temperature control, and prefer JSON‑mode or tool calling for structured outputs that slot directly into your pipeline without brittle parsing.
Mistral 3.1 Small is a AI model from Mistral AI. Pricing is $0.10 per 1M input tokens and $0.30 per 1M output tokens. The context window is roughly 128,000 tokens, allowing long prompts and documents. Common uses include chat assistants, summarization, knowledge search, report drafting, and, when supported, image understanding or tool use. It integrates well into web apps, backends, and automation pipelines where latency and reliability matter.
Mistral 7B by Mistral AI is a capable general‑purpose AI model for chat, writ...
Mistral Medium by Mistral AI is a multimodal AI model for chat, writing, and ...
Mixtral 8x7B by Mistral AI is a capable general‑purpose AI model for chat, wr...
Join only me, who has switched to API.chat and is saving on AI expenses while enjoying a better experience.