Search the GenAI Directory

Traders tracking food and farm prices have to hunt through series and commentary to see what moved. This is a hosted service where you search price reporting content by typing a question, and it also summarises the market notes alongside the numbers. The publisher describes it in the context of AI search & IQ Insights.

Growers and educators want an answer that matches what their local extension service already publishes. This is a web assistant that searches United States extension guides and factsheets and answers questions from that material.

Spotting a leaf problem early is hard when you are looking at hundreds of plants. These are downloadable model weights that sort photographs of leaves into disease categories, ready for developers and researchers to plug into their own tools.

Sorting large piles of agrifood text by topic is slow if you start with nothing. This is a bundle of downloadable text sorting models plus market data files you can reuse inside your own text processing setup.

Comparing crop production, trade or emissions across countries usually starts with hunting for the right table. This is a developer interface to the United Nations food and agriculture statistics collection, so your code can query it directly.

Production figures for crops and animal products across Africa are scattered and awkward to combine. This is a downloadable table of those statistics repackaged so you can pull it straight into a data frame.

You want crop and animal product output numbers for Asian countries without hand cleaning spreadsheets. This is a downloadable version of official production statistics, arranged for direct loading into a model or notebook.

Reporting on the climate footprint of animal farming means finding figures per country. This is a downloadable table of livestock emissions data for Africa, hosted on a model and dataset sharing site.

Carbon accounting across Asian livestock systems stalls without country level emissions numbers. This is a downloadable table of livestock emissions statistics for Asia, published in a format that code can read directly.

Manure figures matter for nutrient and emissions reporting, but they are buried in statistical portals. This is a downloadable table of livestock manure indicators for African countries, shared in a form scripts can read.

You need manure related numbers for Asian countries to support nutrient or emissions work. This is a downloadable table of livestock manure indicators, republished so you can load it without extra cleaning.

Understanding how animal keeping is arranged across African countries takes tidy comparative figures. This is a downloadable table of livestock pattern indicators, republished for use in models and analysis code.

Comparing livestock keeping across Asian countries is easier when the indicators sit in one file. This is a downloadable table of livestock pattern statistics for Asia, shared in a format ready for analysis.

Livestock questions about feed, breeding, health and grazing come up daily on a working property. This is a chat assistant sold alongside animal identification hardware that discusses those topics and offers suggestions.

Stepping back to review how a whole farm is laid out and managed rarely happens amid daily work. This is a copyable prompt that walks a chat assistant through a structured look at fields, crops, livestock and sustainability.

A US and Canada agricultural assistant that uses large language models and retrieval over government sources to answer crop, pesticide-label, nutrient and acreage questions with citations.

A garden robot needs software to read plans, move safely and record what it did. This is the open operating system and control code that runs FarmBot machines, which you can install and change yourself.

You want an AI assistant to update farm records rather than doing it yourself in another tab. This is a hosted plug-in interface for AI assistants that lets them carry out actions in a farm management system.

Farmers want advice that fits their own area and crop, not general guidance. This is a Japanese farming app whose chat function is connected to a research agency’s language model, alongside search and record keeping features.

Smallholder farmers rarely have an adviser on hand and may not read the language the guidance is written in. This is a hosted advisory app that takes questions as text, voice or pictures in several languages and answers from checked farming material.

You want to run your own farm advisory chatbot rather than depend on someone else’s service. This is the open code behind a farming advice assistant, including the parts that load reference material and the chat front end.

You want to build a farmer helpline chatbot that copes with different languages and spoken questions. This is a public code repository with a web front end, a server and a small trained model for farm questions that you can run and adapt.

General farming and rural living questions come up and you want somewhere to talk them over. This is a preconfigured chat assistant, opened through a shared link, covering modern farming practice, sustainable methods and country life.

You want your assistant to answer farm questions the way an experienced grower would, every time. This is an installable add-on for the Claude assistant that carries farming knowledge and working habits along with it.

Setting up a chat assistant to help run a farm means writing all the instructions yourself first. This is a paid marketplace prompt that configures such an assistant, or drafts a farm management plan, once you paste it in.

You have a farm robot and want to add your own sensors or software rather than wait for the maker. This is an open development kit with code and documentation for building extensions on that robot platform.

Machine and field records from different tractors and older systems rarely line up when you try to use them. This is a documented programming interface for pulling farm machine and field data into your own tools.

Agricultural data sits in separate systems that were never meant to talk to each other. This is open source infrastructure for moving and sharing that data between services, sitting behind tools that farmers use.

Indian farmers making day-to-day crop decisions often have no adviser to check with. This is a mobile assistant that answers crop questions in conversation and suggests what to do about them.

An enterprise agronomy command center for North American teams that combines field history, scouting, monitoring, product evidence and grower follow-through with a grounded agronomy agent and agentic workflows.

A hosted MCP gateway that currently connects AI clients to John Deere Operations Center data, weather and a documented field-intelligence engine through normalized tools.

Claims about field boundary detection are hard to compare without shared test material. This is a downloadable dataset of labelled farm parcels from many countries, prepared so different boundary models can be measured on the same ground.

Linking a field boundary to its soil, weather and cropping history usually means stitching several datasets together. This is an open data service covering thousands of fields with those layers already joined for you.

Sometimes you just want field shapes for an area rather than the model that produces them. This is an openly licensed worldwide dataset of field boundary outlines for recent seasons, made from satellite imagery and free to download.

You need field outlines for a region and would rather not trace them yourself. This pairs downloadable model weights and a hosted service for drawing boundaries with an already made worldwide layer of field shapes for recent seasons.

Farm accounts arrive as invoices and processor statements that someone keys in line by line. This is a hosted feature that reads farm financial paperwork, including livestock payment slips from meat processors, and matches it into the farm books. The publisher describes it in the context of incl. kill sheet automation.

Before trusting an AI system on a farming or food task you need a way to check it. Estimates fish species, weight in grams and length in centimetres per segmentation mask, with submitted code run on hidden test data.

Preparing image sets often means trimming away watermarks and clutter one file at a time. This is a code tool that uses an open image understanding model to crop pictures automatically while keeping as much of the original as it can.

Grocery and supplier teams pay for research reports and then cannot find the number they half remember. This is a members’ assistant that searches an association’s own reports, surveys and webinars and links you back to the source.

If you are building a farm advice tool, you may want a model that has already seen agricultural writing. This is an openly released language model for agriculture, with weights and code you can download and run.

Wheat photos taken in the field are messy, and training a vision model on them from nothing needs a lot of labelled data. These are downloadable image model weights pre trained on field wheat pictures, in three sizes, ready to adapt.

Teaching software to recognise a dish from a photo needs plenty of labelled meal pictures. This is a collection of food photographs sorted into 101 dish categories, hosted for straightforward download.

Sometimes you want a small slice of food data to try a prompt or pipeline before scaling up. This is a hosted subset of five hundred food related examples, prepared for use with language models.

General model tests say little about whether an assistant handles food and nutrition questions sensibly. This is a question and answer test set focused on food and nutrition topics, for scoring model responses.

Food and drink leaders are being asked what AI means for growth and do not have an answer yet. This is an advisory engagement covering strategy, operations, skills and demand planning as text generating AI is brought in.

Early planning for a new drink or snack involves a lot of scattered research before anything is decided. This is a hosted assistant service that supports those first product planning steps for food, drink and consumer goods teams.

A food company’s own procedures and hard-won know-how usually live in documents nobody reads or in one person’s head. This is an add-on module that makes that internal material askable and builds guided chats and helpers into product development work.

Small food and drink firms need market and regulatory advice but cannot keep consultants on retainer. This is a mixed service pairing an AI adviser and referenced market reports with the option to book time with human specialists.

This is a locally run MCP server that exposes USDA FoodData Central food search and nutrient data to compatible assistants. It requires a free USDA FoodData Central API key.

You want your assistant to look up real nutrition figures for foods instead of guessing them. This is a small server using a plug-in interface for AI assistants, giving them search access to a public food and nutrient database.

Coming up with a new drink or snack idea and knowing whether it is worth pursuing usually takes weeks of research. This is a hosted assistant for food and drink teams that surfaces trends, drafts product ideas and checks them against its own data.

Assistants often stumble on regional dishes and eating customs outside the mainstream. This is a test set combining photographs and questions about Chinese food culture, used to check fine detail understanding.

Your app has a photo of a meal and needs to know what dish it is. This is a downloadable image model, trained on a public food photo collection, that sorts pictures into a hundred or so dish types.

An ingredient list runs long and you are unsure which additives deserve a closer look. This is a preconfigured chat assistant, opened through a shared link, that points out ingredients and additives worth checking.

Nutrient and allergen details arrive as supplier documents and label pictures, and someone has to retype them. This is a hosted helper that reads those files and drops the values straight into your labelling work.

Supplier certificates, spec sheets and audit reports pile up as files that someone has to retype before an audit. This is a team you hire to build a reading pipeline with human checks that files the extracted fields into your compliance system.

If you are building something that handles food information, you need a way for the assistant to fetch it. This is an openly available plug-in interface offering a handful of food data tools for AI assistants to call.

Building anything that talks about calories or nutrients needs a table of food composition to work from. This is a publicly shared nutrition dataset you can download and use for training, testing or reference.

Answering food questions from a linked database of facts is a different skill from free text answering. This is a public challenge with roughly ten thousand food questions and answers built over a structured food fact store, plus code.

A food safety plan needs hazard analysis, control points, flow charts and written procedures before an inspector will sign off. This is a hosted workflow that drafts those parts of the plan for you to review and finish.

Tracking what gets recalled and why means reading notices from many different national authorities. This is a collection of food recall announcements with expert labels for the hazard involved and the product concerned, in several languages.

Classifying recall notices by product and hazard is repetitive reading that teams would rather automate. This is a dataset of recall texts together with code for training and comparing classifiers on the task.

Logging a meal usually means guessing portions and typing everything in by hand. These are downloadable model weights and code that recognise foods in a photo, estimate portion size and work out calories across about a hundred food categories.

Before trusting an AI system on a farming or food task you need a way to check it. Detects individual food items in real meal photographs, scored with COCO average precision and average recall.

Food safety teams wade through long documents and need short summaries they can trust. This is a published benchmark for summarising food safety documents, with source texts and reference summaries for scoring.

Earning a food safety certification and holding onto it means matching daily practice to an audited standard. This is a preconfigured chat assistant, shared by link, answering questions about globally recognised food safety standards and certification.

Rules differ by site, and someone still has to turn them into checklists that staff will actually use. This is a preconfigured chat assistant, opened through a shared link, that discusses food safety rules and drafts tailored checklists.

Temperature logs, inspection notes and corrective actions pile up on paper and get chased by hand. This is a collection of ready-made automation recipes for a popular workflow tool, covering food safety records, audits and follow-up tasks.

Staff ask food safety questions and you would rather not repeat the same explanations every week. This is a free set of instructions you paste into a common chat assistant so it answers food safety questions in detail.

Judging the risk in a food supply chain means checking recalls, inspection ratings, trade flows, air quality and weather separately. This is a plug-in interface, with open code, that gathers seven public sources and scores them into a single written risk report.

Menus, labels and recipes lose meaning when translated by a general tool. This is a downloadable set of matched Korean, English and Chinese food texts for building translation models in that subject area.

If you are building something food related, sending every request to a hosted model is not always practical. These are downloadable model weights, with code, for a language model focused on food topics that you can run locally.

Finding a plant ingredient with a health story behind it means wading through scientific papers. This is a hosted platform that searches research on plant compounds and their reported health effects and points to ingredients worth a closer look.

People ask nutrition questions constantly and a dietitian is not always available. These are downloadable adapter weights, with code, that tune an existing language model to answer questions about food and nutrition.

Working towards a food safety certification involves paperwork, evidence and repeated audits. This is a preconfigured chat assistant, shared by link, that answers questions about achieving and maintaining FSSC 22000 certification.

Getting boundary data downloaded, models trained and outputs turned into map shapes involves a lot of glue code. This is an installable command line tool and code library covering those steps for one open field boundary project.

Input dealers field the same order, price and stock questions all day long. This is a hosted AI agent for farm input resellers and distributors that handles those routine exchanges and connects into their existing systems.

A 270M-parameter Gemma model specialized for function calling. It is designed to be fine-tuned for a defined task so a small local agent can turn natural-language requests into structured API actions. It is not agrifood-specific and task-specific fine-tuning and testing are expected.

Agricultural research reports are hard for software to read in their original form. This is a downloadable collection of farm research documents converted into a consistent text format for grounding or training a model.

For quick farming questions, many people reach for whatever general assistant they already have access to. This is a general purpose language model, called through a programming interface, that has been measured on a set of agricultural advice questions.

A general-purpose multimodal Gemma variant built for everyday devices. It processes text, images and audio and uses parameter-efficient techniques intended to reduce local memory and compute requirements. It is a foundation component, not an agrifood-tuned model.

A general-purpose, open-weight multimodal foundation-model family from Google DeepMind. Gemma 4 accepts text and images, with audio on selected sizes, supports reasoning and function calling, and includes E2B and E4B variants designed for local execution on phones and laptops. It is not trained specifically for agrifood.

You need a model that responds to farming instructions without you assembling a training set first. These are downloadable model weights, with code, for an instruction following language model given extra agricultural examples.

Pulling a report out of a food company business system usually means waiting on somebody technical. This is a hosted feature where you type a question in ordinary language and get figures or a report back without writing code.

Before trusting an AI system on a farming or food task you need a way to check it. Predicting maize hybrid grain yield across untested environments, with entries scored on held-out data through an EvalAI leaderboard.

Before trusting an AI system on a farming or food task you need a way to check it. Produces 10 metre cropland extent maps for test regions, assessed for accuracy against test samples and practicality.

Before trusting an AI system on a farming or food task you need a way to check it. Earth observation benchmark for pretrained models, including crop type and cashew plantation segmentation tasks, downloaded from Hugging Face.

Before trusting an AI system on a farming or food task you need a way to check it. 31 geospatial vision-language tasks with over 10,000 verified instructions, including crop type classification and farm pond change detection.

Several satellite models claim to handle crop mapping, and comparing them yourself takes weeks. This is a test setup plus already adapted model files for measuring those models on agricultural region labelling and change spotting.

Writing a fertiliser prescription means balancing nutrient targets, soil limits, buffer zones and local rules, then getting the file into the tractor. This is a hosted assistant that walks through those checks, shows its working and exports the finished plan to machinery. The publisher describes it in the context of reasoning agronomy assistant.

A field’s soil tests, yield history, boundaries and application records usually sit in a platform you have to click through. This is a hosted plug-in interface that hands that farm context to an AI assistant, plus tools for maps, sampling plans and machine exports.

Sales notes, invoices, contracts and inspection forms pile up in an agricultural business and get typed in twice. This is a set of hosted assistants you configure for those tasks, pulling details from paperwork and moving them through your process.

IFPA describes the Global Intelligence Engine as a members-only platform that produces tailored insights from produce, floral, consumer and retail data. This listing records the publisher’s description; no independent product test was performed.

Planning around meat supply in Africa needs figures broken out by animal type over time. This is a downloadable table of meat production series by country and livestock type, ready to load into code.

Before trusting an AI system on a farming or food task you need a way to check it. Reviews risk assessment, safety function design and validation plans of autonomous farm robots against ISO 18497, ISO 25119 and Machinery Regulation 2023/1230.

Before trusting an AI system on a farming or food task you need a way to check it. Localises wheat heads in field images and measures robustness using average domain accuracy across countries and sensors.

Before trusting an AI system on a farming or food task you need a way to check it. Segments wheat plant parts including leaves and stems in field images spanning wheat types, growing regions and growth stages.

Training a model to count wheat heads needs field photos from more than one farm. This is a downloadable image collection gathered across many regions, with wheat heads marked up so you can train or score detectors.

Formulating a weedkiller for particular weeds and crops involves trade-offs you may want to talk through. This is a preconfigured chat assistant, opened through a shared link, focused on glyphosate based herbicide formulation and adjustment.

Advisory teams often want to know how far a general assistant gets with crop questions. This is a general purpose language model, available through a paid programming interface, that was included in an agricultural advisory evaluation.