Command A
Neural Network
Command A for code and agentic scenarios — Cohere's open-weights model with a 256k token context.
Max answer length
(in tokens)
Context size
(in tokens)
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works Command A?
Command A is an open-weights model by Cohere with 111 billion parameters and a 256k token context window. It is built around three pillars: agentic systems, multilingual tasks, and coding, allowing it to confidently handle long scenarios with tools and analyze entire repositories. Open weights mean architectural transparency, and the large context means documentation, logs, and source code can be provided without splitting them into small fragments. In BotHub, you connect to Command A without a VPN or foreign card: pay with Russian cards based on usage, no subscription, and unused tokens do not expire. Alongside it in one window are over 250 models, making it easy to switch and compare answers, while a unified API allows you to change models in a multi-model system without rewriting integrations; requests are encrypted and not saved, and companies have access to contracts, invoices, electronic document management, and an admin panel with limits. Developers use it for refactoring and debugging, analysts for parsing long reports, product teams for building agents, and support teams for multilingual ticket processing.Frequently asked questions about Command A
You can use generated results for commercial purposes. You own all rights to the content you create. The only restriction: make sure your prompt does not include copyrighted third-party material. You are responsible for respecting the rights to any input data.
Command A by Cohere is a text model that accepts text, documents, and images as input. The 256,000 token context window allows you to upload a voluminous report or an entire codebase, making the model suitable for contract analysis, data extraction from scans, summarization, editing, and coding.
Up to 8,192 tokens per response, which is roughly several thousand words: a long article, a documentation section, or a large code module. If you need more volume, break the task into parts: the 256,000 token context allows you to keep the entire previous text in the dialogue.
Our data does not indicate a separate reasoning mode for Command A, so the model answers immediately. You can get a step-by-step breakdown via a prompt: ask it to write out the logic step-by-step. If you need a built-in chain of thought, switch to a reasoning model in the same window in BotHub.
Our data does not mention function calling or strict JSON output, so we cannot promise it in advance — please check with your scenario. You can request a JSON response directly in the prompt. Through BotHub's unified OpenAI-compatible API, you can switch models without rewriting your integration.
Images and documents — yes: upload screenshots, diagrams, scans, PDFs, and ask it to analyze the content or extract data. Audio and video input is not noted in our data; for sound and clips, BotHub has separate models available in the same interface.
We do not have data on languages, so we cannot provide an assessment — it is better to run your typical query and compare the result with another model, which can be done in one window. The service itself is Russian-speaking: access from the Russian Federation without a VPN or foreign card, payment in rubles.