Opening article
We're thrilled to share that Cohere is now a supported Inference Provider on HF Hub! Starting now, you can run serverless inference to the following models via Cohere and Inference Providers: You can use Cohere models directly on the Hub either on the website UI or via the client SDKs. You can search for Cohere models by filtering by the inference provider in the model hub .
We're thrilled to share that Cohere is now a supported Inference Provider on HF Hub! This also marks the first model creator to share and serve their models directly on the Hub.
Cohere is committed to building and serving models purpose-built for enterprise use-cases. Their comprehensive suite of secure AI solutions, from cutting-edge Generative AI to powerful Embeddings and Ranking models, are designed to tackle real-world business challenges. Additionally, Cohere Labs , Cohere’s in house research lab, supports fundamental research and seeks to change the spaces where research happens.
Starting now, you can run serverless inference to the following models via Cohere and Inference Providers:
32-billion parameter model with advanced capabilities optimized for a variety of vision-language use cases, including OCR, captioning, visual reasoning, summarization, question answering, code, and more. It expands multimodal capabilities to 23 languages spoken by over half the world's population.
Use the huggingface_hub python library to call Cohere endpoints by defining the provider parameter.
Ideal for low-cost or low-latency use cases, bringing state-of-the-art performance in its class of open-weight models across real-world tasks. This model offers a context length of 128k. It delivers a powerful combination of multilingual support, citation-verified retrieval-augmented generation (RAG), reasoning, tool use, and agentic behavior. Also supports 23 languages.
Optimized for demanding enterprises that require fast, secure, and high-quality AI. Its 256k context length (2x most leading models) can handle much longer enterprise documents. Other key features include Cohere’s advanced retrieval-augmented generation (RAG) with verifiable citations, agentic tool use, enterprise-grade security, and strong multilingual performance (support for 23 languages).
Focuses on state-of-the-art multilingual support, applying the latest research on multilingual pre-training. Supports Arabic, Chinese (simplified & traditional), Czech, Dutch, English, French, German, Greek, Hebrew, Hindi, Indonesian, Italian, Japanese, Korean, Persian, Polish, Portuguese, Romanian, Russian, Spanish, Turkish, Ukrainian, and Vietnamese with 128K context length.
Cohere and Cohere Labs bring a swathe of their models to Inference Providers that excel at specific business applications. Let’s explore some in detail.
You can use Cohere models directly on the Hub either on the website UI or via the client SDKs.
You can find all the examples mentioned in this section on the Cohere documentation page .
Cohere on Hugging Face Inference Providers 🔥. Covered by Hugging Face on April 16, 2025.