Blog
Explore our latest insights
Delve into our recent articles, offering a diverse range of topics, expert perspectives, and thought-provoking ideas across various fields.

Plant.health major expansion: from 90 to 500+ diseases
We’re rolling out a major upgrade to the Plant.Health disease detection model—built on a new architecture, trained on a much larger dataset, and backed by an expanded, multilingual knowledge base. The model now distinguishes 548 plant-health classes (up from 90) across insects, fungi, bacteria, and abiotic stresses—while maintaining overall accuracy.
Introducing open-source Kindwise Router Models
We’ve open-sourced the Kindwise Router — a lightweight image model that helps your app automatically decide which Kindwise API to use (Plant.id, Insect.id, Crop.health, etc.) right on the device, before anything is uploaded.

New plant.health feature: follow-up questions
With this new feature, users can receive automated, targeted, easy-to-answer, expert-backed questions to help refine potential issues and guide plant care treatments.

Using LLMs for Descriptions of Plants, Insects, and Fungi
We tested LLMs to improve multilingual descriptions of plants, insects, and fungi where Wikipedia entries were inadequate. GPT-4o outperformed smaller models such as GPT-4o mini and Mistral Pixtral 12B, especially in grammar and factual accuracy. It performed better in widely spoken languages and common species. In the end, GPT-4o was used to generate descriptions in 28 languages, providing better coverage for the Kindwise API.
.png)
Production release: insect.id & mushroom.id
Explore the exciting news and improvements in insect.id and mushroom.id coming out with production release.
Our most popular solutions
.jpg)

.webp)






b.jpg)






