Solar Mini 4
Neural Network
Solar Mini 4 is a compact Upstage language model with 524K context, built for agentic scenarios; available via BotHub API.
Specifications
- Input
- text, documents
- Output
- text
- Context size
- 524 288 tokens
- Max answer length
- 131 072 tokens
- Reasoning effort
- none, minimal, low, medium, high, extra high, max
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works Solar Mini 4?
Solar Mini 4 is an Upstage language model built on a mixture-of-experts architecture: 35 billion parameters, with about 3 billion active. This design makes the model efficient and fast, while the 524K token context window allows handling large documents, long dialogues, and agent step chains in a single request. Its primary profile is agentic scenarios where response speed and reasoning quality are crucial. In BotHub, you connect to Solar Mini 4 without VPN or foreign cards: pay with Russian cards in rubles based on usage; tokens do not expire. Access over 250 models in one window to easily switch and compare responses. A unified OpenAI-compatible API lets you swap models in multi-agent systems without rewriting integrations. Data is encrypted; companies get contracts, invoices, EDI, and an admin panel with limits. Useful for building autonomous agents and tool-based pipelines, analyzing long reports and entire threads, batch processing documents, creating responsive support assistants, and prototyping solutions with cost control.Frequently asked questions about Solar Mini 4
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.
solar-mini4 by upstage is a text model with reasoning mode, document handling, and function calling. It is suitable for analyzing long materials, writing and refactoring code, STEM tasks, structured data extraction, and embedding into your own services via the unified BotHub API.
Up to 131,072 tokens in a single response — enough for a voluminous document, a large code module, or a detailed breakdown. The context window is 524,288 tokens, so long output can be continued in the same dialogue without losing initial task details.
Yes, reasoning before answering is confirmed. The model breaks down the task into steps, checks intermediate conclusions, and only then formulates the result. This helps with math, logic, code debugging, and multi-step questions where the path to the result is as important as the result itself.
Yes. The model supports function calling and structured JSON output, making it easy to connect to tools, databases, and internal services. In BotHub, this is available via a unified OpenAI-compatible API: you don't need to rewrite integrations when switching to another model.
Besides text, the model accepts documents: you can upload a file and ask for a summary, key points, or answers based strictly on the content. Support for images, audio, and video is not indicated in our data. If you need such formats, switch to a multimodal model in the same BotHub window.
We do not have data on prompt and response languages, so we cannot say for sure. It is easier to test it with your own task: open solar-mini4 in BotHub, pay with a Russian card in rubles without a VPN, and compare the output with other models in one window.