Introducing Command A+: Making sovereign agentic capabilities available to all
Blog post from Cohere
Cohere has released Command A+, an open-source mixture-of-experts (MoE) model designed for high-performance agentic tasks with minimal compute overhead, under an Apache 2.0 license. Command A+ integrates and enhances the capabilities of previous Command series models, expanding multilingual support from 23 to 48 languages and improving performance in reasoning, multimodal understanding, and coding tasks. With a focus on efficiency, it offers substantial improvements in speed and latency, making it possible to run on limited hardware like two NVIDIA H100s or a single NVIDIA Blackwell GPU without compromising quality. Available for download on Hugging Face in various quantizations, Command A+ is optimized for developer-focused use, enabling practical and scalable AI deployment. This model represents a significant step in Cohere's mission to make sovereign AI a reality, providing enterprise-grade capabilities that can be locally deployed and controlled by developers, thus contributing to the broader goal of AI independence.
| Trend | Post Mentions | Total Month Mentions | Posts | Companies | MoM |
|---|---|---|---|---|---|
| LLM | 3 | 10,547 | 1,921 | 254 | +37% |
| Secrets Management | 2 | 2,462 | 459 | 124 | +13% |
| AI Agents | 1 | 6,625 | 1,634 | 282 | -6% |
| Agent Evaluations | 1 | 187 | 70 | 28 | -31% |
| MCP | 1 | 13,294 | 944 | 215 | +3% |
| RAG | 1 | 2,384 | 400 | 99 | +78% |
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.