Unofficial AI-summarized news site (not affiliated with any AI company)
AI News JP / www.ai-news.jp
🟠 Important AI Summary · Source: Mistral AI

New AI endpoints are now available in early access, offering a variety of options!

Mistral AI Launches New AI Endpoints in Early Access

Original: La Plateforme | Mistral AI

Importance: 新しいAIエンドポイントのリリースは多くのユーザーに影響を与えるため。

Summary

Mistral AI is offering new AI endpoints in early access. There are three chat endpoints and one embedding endpoint, each with different performance and price trade-offs. Notably, Mistral-tiny is cost-effective and serves only English with a score of 7.6. Mistral-small supports multiple languages with a score of 8.3, while Mistral-medium is a prototype model boasting a score of 8.6. An API is also provided, with services gradually expanding.

Key Points

  • New AI endpoints available in early access
  • mistral-tiny is English-only with a score of 7.6
  • mistral-small supports multiple languages with a score of 8.3
  • mistral-medium is a prototype with a score of 8.6
  • API provides Python and JavaScript clients
View developer notes (APIs, breaking changes, migration)

Mistral AI is launching three chat endpoints (mistral-tiny, mistral-small, mistral-medium) and one embedding endpoint (Mistral-embed) in early access. Mistral-tiny utilizes Mistral 7B Instruct v0.2, serves only in English, and scores 7.6 on MT-Bench. Mistral-small features the Mixtral 8x7B model with a score of 8.3, while Mistral-medium serves a prototype model with a score of 8.6. The embedding endpoint employs a 1024-dimensional embedding model, achieving a score of 55.26 on MTEB. The API provides Python and JavaScript client libraries, allowing users to set system prompts.

モデルAPI/SDKパフォーマンスAudience: 一般ユーザーAudience: 開発者

Source: https://mistral.ai/news/la-plateforme

Outlet: Mistral AI

This article is an AI-generated summary (OpenAI GPT-4o-mini) of publicly available information from Anthropic, OpenAI, Google, Meta, Mistral, DeepSeek, Sakana, and other vendors. The original source URL is always provided in accordance with fair-use citation requirements. Summaries are AI-generated and may contain mistranslations or misinterpretations. Always verify details with the original source.