Mistral: Mistral Embed
mistral-embed- Input / Output
- $0.10 / $0.00
- Context
- 8K
- Modalities
- Embedding
About Mistral Embed
Mistral Embed is Mistral AI's general-purpose text embedding model. It converts sentences, paragraphs, and documents into vectors for semantic search, clustering, classification, and the retrieval step of retrieval-augmented generation. It is the text-oriented sibling of Codestral Embed, which Mistral built for source code. Choose it when your corpus is prose such as help articles, contracts, or notes, and you want a separate vector model rather than a chat call.
Where it works well
- General text coverage, so one model can index help-center articles, reports, and chat transcripts together.
- Fits the retrieval step of retrieval-augmented generation, where a question is matched to the passages most likely to answer it.
- Supports clustering and document classification from the same vectors, without a separate model per task.
- Part of Mistral's embedding line, so code and prose can be indexed with two models from one maker.
When to choose another model
- For source-code retrieval, Mistral's code-specific Codestral Embed is the model built for that task.
- It returns vectors only; it cannot answer questions or summarise the passages it retrieves.
- Inputs are capped at a modest token length, so long documents must be split into chunks before embedding.
Getting started
Create API key
Create a key in Console, then use it with every model on the platform.
Send your first request
Copy the example for your language and run it against the endpoint.
EmbeddingTokenLab endpointPOST/v1/embeddingscurl -X POST "https://api.tokenlab.sh/v1/embeddings" \ -H "Authorization: Bearer sk-xxx" \ -H "Content-Type: application/json" \ -d '{ "encoding_format": "float", "model": "mistral-embed", "input": "The quick brown fox jumps over the lazy dog." }'
Parameters
Embedding
| Parameter | Values and limits | Default | Notes |
|---|---|---|---|
encoding_format | float, base64 | Not provided | Not provided |
input | Not provided | Not provided | Not provided |
Pricing
The Verified price applies to Verified, which costs less on most models. The Official price is the model maker's published price and applies to the more reliable Official route. Auto bills the route that completes the request.
Embedding
per 1M tokens- Official price
- Input $0.10 / Output $0.00
- Verified price
- —
- Discount
- —
| Official priceper 1M tokens | Verified priceper 1M tokens | Discount | |
|---|---|---|---|
| Embedding | Input $0.10 / Output $0.00 | — | — |
Usage & activity
Success rate is the share of requests that completed. Latency is how long a full response takes; P95 means 95% of requests finished within that time.
Usage & availability
Last 24 hours- Requests
- Success rate
- P95 latency
- Total tokens
Data is based on aggregate user requests, excluding status checks.
Open in Console
Open Mistral Embed in Console with a prompt ready to edit or send.
Help me try mistral-embed with embedding at /v1/embeddings. Show the result and cost.
Use cases
Help-center semantic search
Embed every article once, embed each visitor question at query time, and show the nearest articles even when no keywords overlap.
Retrieval for question answering
Fetch the top passages from internal documents and pass them to a chat model so its answer is grounded in your own material.
Topic clustering
Group thousands of survey replies or tickets by meaning to see which themes recur, without defining categories in advance.
Document classification
Train a light classifier on the vectors to send incoming documents to the right team or label.
Prompt examples
How do I change the billing contact on my account?
Quarterly report: churn rose in the mid-market segment after the pricing change in March.
The delivery driver left the parcel at the wrong address and support has not replied.
This model has conditional pricing. Monthly token totals alone cannot produce a reliable estimate; use the detailed pricing for the request specification, cache, and applicable time window.
FAQ
What is Mistral Embed for?
Turning text into vectors you can compare. Typical uses are semantic search over documents, grouping similar texts, classifying documents, and retrieving passages for a chat model to answer from. It does not generate text.
Should I use Mistral Embed or Codestral Embed?
Use Mistral Embed for natural-language content and Codestral Embed for source code and code-related queries. If your corpus mixes both, index each part with its own model and keep the vectors in separate collections, since vectors from different models are not comparable.
Can I mix vectors from different embedding models?
No. Vectors are only comparable when they come from the same model. If you switch from Mistral Embed to another embedder, re-embed the whole corpus and rebuild the index, including any stored queries.
How do I handle long documents?
Split them into passages before embedding, because each input has a token limit. Paragraph or section-sized chunks with a little overlap usually retrieve better than whole documents, and you can store the document title with every chunk.
Does it work through the OpenAI-style embeddings request?
Yes. You send a list of strings to the embeddings endpoint and receive one vector per string, returned in the same order as the input.
How much does Mistral Embed cost?
On TokenLab, Mistral Embed costs Input $0.10 / Output $0.00 per 1M tokens. The pricing table above shows the full breakdown. Rates depend on the billing unit, specification, and usage. Compare matching conditions in the model's detailed pricing; a single rate does not determine the total cost.
Which endpoint should Mistral Embed use?
Use https://api.tokenlab.sh/v1/embeddings for Mistral Embed. The request example below shows the matching code shape.
Which operations does Mistral Embed support?
Mistral Embed supports Embedding. Select an operation above to see its endpoint and request example.
Compare Mistral Embed
Sources
Reviewed Oct 2, 2026
