Cohere Command A+: a 218B-parameter model built for agents
Cohere has released Command A+, an open model with 218 billion parameters designed for agentic workflows. Thanks to W4A4 quantization, it runs on two H100 GPUs and supports 48 languages. It is also the first in the line to gain multimodal reasoning capabilities. Command A+ merges the four variants of the previous Command A series into a single, powerful general-purpose model.
AI-processed from MarkTechPost; edited by Hamidun News
Cohere has released Command A+, a next-generation open model built for agentic workflows.
How many GPUs does Command A+ need?
Thanks to W4A4 quantization, the model runs on two H100 GPUs. Competing models with 300B+ parameters require eight to sixteen GPUs.
Which languages does Cohere Command A+ support?
The model supports 48 languages.
How many GPUs does Cohere Command A+ run on?
Two H100 GPUs, thanks to W4A4 quantization, which stores both weights and activations in 4-bit precision. Competing models with 300B+ parameters require eight to sixteen GPUs, making them four times more expensive to run.
How many parameters does Cohere Command A+ have?
218 billion parameters in a Sparse Mixture-of-Experts architecture. The model merges the four previous Command A variants into a single general-purpose model.
- Nov 1, 2026Google's StreamVibe video agent beats other models by 10.7% on response timing and quality
- Nov 1, 2026New test catches hidden AI agent collusion by measuring how a partner reacts to a paraphrase
- Nov 1, 2026Illinois compiled 12 151 open problems and tested whether AI can solve them
- Oct 31, 2026AI engineer writes a single reward function for agent teams: win rate climbs from 51% to 90%
Key Questions Answered
How many GPUs does Command A+ need?
Thanks to W4A4 quantization, the model runs on two H100 GPUs. Competing models with 300B+ parameters require eight to sixteen GPUs.
How many GPUs does Cohere Command A+ run on?
Two H100 GPUs, thanks to W4A4 quantization, which stores both weights and activations in 4-bit precision. Competing models with 300B+ parameters require eight to sixteen GPUs, making them four times more expensive to run.
How many parameters does Cohere Command A+ have?
218 billion parameters in a Sparse Mixture-of-Experts architecture. The model merges the four previous Command A variants into a single general-purpose model.
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.