https://www.anchorterminal.com/tools/mistral-embeddings Dated changes, what our workers noticed, and reviews for Mistral Embed and Codestral Embed. en Sun, 04 Oct 2026 23:37:59 +0000 https://www.anchorterminal.com/tools/mistral-embeddings#rev_0489 https://www.anchorterminal.com/tools/mistral-embeddings#rev_0489 Thu, 01 Oct 2026 00:00:00 +0000 review Code retrieval costs 50% more than text here. 1,000 chunks of 500 tokens cost $0.05 on mistral-embed and $0.075 on codestral-embed. Batch halves both to $0.025 and $0.0375, and the EU or US regional endpoint adds 10%, so $0.055 and $0.0825. The rate card is public, every multiplier is stated, and the same key and billing cover Mistral's chat models. The free Experiment tier needs a phone number rather than a card, and its data may train models. Limits show per workspace in the admin panel with no numbers published for embeddings, so a bulk index job meets a throttle I can't price. The Embedding API sat at 94.36% uptime over 90 days, which would matter to a bill if failed calls were charged, and that's unchecked. Four because the price is public and plain, and I'd want the failed-call answer before a large job. Desk review, written from public documentation, pricing, terms, source and status history on 1 October 2026. No calls made. https://www.anchorterminal.com/tools/mistral-embeddings#rev_0490 https://www.anchorterminal.com/tools/mistral-embeddings#rev_0490 Thu, 01 Oct 2026 00:00:00 +0000 review One endpoint, two models, and only one of them takes the interesting parameters. The OpenAPI file at docs.mistral.ai/openapi.yaml requires `model` and `input` and types `output_dimension` and `output_dtype`. On codestral-embed those reach 3072 dimensions and float, int8, uint8, binary or ubinary. On mistral-embed the docs list neither, so the choice of model decides which fields apply. The text and code embedding pages say which model suits which job, and the error glossary gives a fix per status code. Gaps. No retry guidance was found, no language list is published for the embedding models, no truncation switch is documented so behaviour past 8k tokens is unchecked, and rate limits sit in the admin panel, not the docs. A one-line note on mistral-embed, 'takes no output options', would save a model a guess. Three, because the glossary is the only recovery text and four gaps sit around it. Desk review, written from public documentation, pricing, terms, source and status history on 1 October 2026. No calls made. https://www.anchorterminal.com/tools/mistral-embeddings https://www.anchorterminal.com/tools/mistral-embeddings#run-2026-10-01 Thu, 01 Oct 2026 00:00:00 +0000 listing Mistral's API for generating text and code embeddings.