AI News HubLIVE
In-site rewrite2 min read

Command A+

Cohere launches Command A+, its fastest and most powerful open-source model yet, on Product Hunt. With a 218B MoE architecture (25B active parameters), it supports reasoning, tool use, vision, long context, and multilingual workflows, and can run on just two H100 GPUs. It's designed as an enterprise workhorse for high-performance agents.

SourceProduct Hunt AIAuthor: Zac Zuo

Build incredible products with world-class language AI | Cohere | Product Hunt

Build incredible products with world-class language AI

4.9•13 reviews•

661 followers

Build incredible products with world-class language AI

4.9•13 reviews•

661 followers

Visit website

LLMs

AI Infrastructure Tools

We build high performance, secure language models for the enterprise. Our customizable, high-performance language models work on public, private, or hybrid clouds to ensure data security & exceptional support.

Overview

Launches7

Reviews13

Alternatives

Customers

Team

Awards

More

This is the 7th launch from Cohere. View more

Command A+

Launching today

Cohere’s open enterprise workhorse

Cohere’s fastest and most powerful language model yet: an open-source model for running high-performance enterprise agents with maximum efficiency.

Free

Launch tags:Open Source•Artificial Intelligence•Development

Launch Team

Subscribe

SocialX

Previous Cohere Launches

Tiny AyaLocal, open-weight AI designed for real-world languages

Launched on April 5th, 2026

Cohere TranscribeNew state-of-the-art in open source speech recognition

Launched on March 28th, 2026

Command A ReasoningEnterprise-grade control for AI agents

Launched on August 25th, 2025

Command ACohere's most performant model for the enterprise

Launched on March 17th, 2025

View all Cohere launches

4.9

Based on 13 reviews

Review Cohere?

Reviewers describe Cohere as a quiet, reliable option rather than a flashy favorite. The lone user review calls it an underrated workhorse, especially strong for instruction-heavy workflows and high-volume prompting. Founder reviews from makers of products like Liveblocks, Stack Overflow, and CustomGPT are uniformly positive, but they add no concrete implementation details. With feedback this sparse and mostly one-sided, the clearest takeaway is steady utility and solid throughput.

+10

Summarized with AI

Pros

Cons

Reviews

All Reviews

Most Informative

Pros

boosting search precision (2)

multi-modal embeddings (2)

Turn scattered DMs into contracts you can control and prove.

Promoted

Hunter

📌

Hi everyone!

Apache 2.0 on Command A+ is a big deal!

This is @Cohere’s fastest and most powerful model yet. It uses a 218B MoE architecture with 25B active parameters, built for enterprise agents that need reasoning, tool use, vision, long context, and multilingual workflows.

Command A+ can run on as little as two H100 GPUs with W4A4 quantization, which is pretty impressive considering its overall performance across agentic, multimodal, and multilingual tasks.

Love how they put it:

We want to give developers direct access to enterprise-grade agentic capabilities from experimentation to production.

Kudos!

Report

17h ago

used Cohere to build CustomGPT.ai Researcher

•6 reviews

We used Cohere multi-modal embeddings to select the images for the articles. We found that this works better than GPT-Vision to select high-quality images that are contextually relevant to the surrounding text. This was a hard problem to solve, and Cohere's multi-modal embedding quality helped improve the quality of the selected images.

What's great

multi-modal embeddings (2)high-quality image selection (1)

Report

62 views1yr ago

used Cohere to build Agora

•14 reviews

We are blown away by Cohere Embed 4’s ability to accurately surface relevant products to search queries. E-commerce data is complex, containing images and multifaceted text descriptions. Being able to represent our products in a unified embedding makes our search faster and our internal tooling more efficient.

What's great

boosting search precision (2)unified embedding (1)

Report

50 views12mo ago

•9 reviews

Very underrated workhorse of a large language model. Particularly useful for more instructional workflows. supports high volume prompting with ease.

What's great

large language model (1)instructional workflows (1)high volume prompting (1)

Report

28 views1yr ago