Aion 3.0
Neural Network
AionLabs multi-model system based on GLM — Aion 3.0 for roleplay and narrative storytelling
Max answer length
(in tokens)
Context size
(in tokens)
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works Aion 3.0?
Aion 3.0 is an AionLabs development, not a single neural network, but a bundle of several specialized GLM family models that prepare responses together. Each handles a part of the generation, so the system is built specifically for roleplay and long stories where character personalities and plot development are important. Through BotHub, the model is accessible without a VPN or foreign card: you can pay with Russian cards in rubles, paying only for consumed tokens that do not expire. Over 250 models are available in the same window, making it easy to rewrite the same scene with another neural network and compare results. A unified OpenAI-compatible API allows you to connect Aion 3.0 to your product and switch models without rewriting the integration; chats are encrypted, and data is not saved. In practice, you can run text-based roleplay campaigns with persistent characters, write lines and NPC personalities for games and visual novels, draft plot branches and chapters, bring chatbot and interactive story characters to life, run one episode in several variations, and keep the best one.Frequently asked questions about Aion 3.0
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.
aion-3.0 is a text model from aion-labs with a 131,072 token context. It analyzes long documents, works with code (refactoring, debugging, logic explanation), solves reasoning and STEM tasks, and accepts images and documents as input. Suitable for analytics, development support, and complex multi-step queries.
In a single response, aion-3.0 generates up to 32,768 tokens — that's dozens of pages of text. Enough for a large documentation section, a detailed analysis, or an entire voluminous code file. If there is more material, break the task into parts: the 131,072 token context allows you to keep the entire conversation in one dialogue.
Yes, reasoning before answering is supported. The model builds a chain of thought, breaks the task into steps, and checks intermediate conclusions before providing the final result. This helps in mathematics, logic, analyzing third-party code, and situations where accurate sequence of actions is important rather than a quick guess.
Yes. aion-3.0 supports function calling and structured JSON output, making it easy to integrate into agents and pipelines: the model decides when to call your tool and provides a predictable schema. In BotHub, this works via a unified OpenAI-compatible API.
Images and documents — yes: send a screenshot, diagram, chart, or file and ask to analyze the content, summarize data, or find an error. Audio and video are not listed among the supported input formats. If you need to work with audio or video, you can switch to a specialized model in the same BotHub window.
We do not have data on this model's language capabilities, so we cannot promise specific quality for Russian — it is more reliable to test it on your own tasks. The BotHub interface is in Russian, access from the Russian Federation is possible without a VPN or foreign card, and payment is made with Russian cards based on actual token usage, so testing will take a couple of minutes.