Claude Sonnet 4.5 Batch
Neural Network
Claude Sonnet 4.5 is an Anthropic model for coding and agentic scenarios; Claude API in batch processing mode.
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 1 000 000 tokens
- Max answer length
- 64 000 tokens
- Reasoning effort
- low, medium, high
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works Claude Sonnet 4.5 Batch?
Claude Sonnet 4.5 is an Anthropic model designed for agentic scenarios and development workflows. It performs exceptionally well on engineering benchmarks like SWE-bench Verified, where understanding a repository and completing a task is crucial, rather than just writing a function. The batch variant is designed for batch processing: you send an array of requests and retrieve the results without maintaining an interactive session. In BotHub, the model works from Russia without a VPN or foreign card; payment is in rubles based on actual token usage, and tokens do not expire. Requests are sent via a unified OpenAI-compatible API, allowing you to switch to any of the 250+ models without rewriting your integration. Data is encrypted via AES-GCM, and companies have access to contracts, invoices, and an admin panel with limits. It is useful for mass code reviews, ticket processing, refactoring legacy code, building agents with function calling, batch analysis of documentation and logs, or preparing technical texts from accumulated materials.Frequently asked questions about Claude Sonnet 4.5 Batch
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.
This is an Anthropic text model with a one-million token context: it analyzes large codebases, documents, and entire conversations. It is suitable for writing and refactoring code, debugging, analytics, working with documents and images, as well as for agentic scenarios with tool calling.
Up to 64,000 tokens in a single response — enough for a large code module, detailed technical analysis, or a long document without splitting it into parts. If the task is larger, break it down into steps: the one-million token context allows you to keep the entire history in one dialogue.
Yes, reasoning mode is supported: the model first outlines the solution path and then provides the answer. This significantly helps in mathematics, logic, code analysis, and multi-step tasks where it is important to arrive at the result sequentially rather than guessing.
Yes, the model supports function calling and structured JSON output. You describe a set of tools, and it decides which one to call and with what arguments. Through the unified OpenAI-compatible BotHub API, it is easy to integrate into agents and internal services.
Images and documents are accepted as input, including natively — the model reads screenshots, diagrams, tables, and PDFs and answers based on their content. Audio and video are not supported. The model generates text; it does not create images or sound.
Anthropic models work confidently with Russian: they understand complex phrasing, terminology, and task context. In BotHub, you access it in the usual way via the website or Telegram bot, without a VPN or foreign card, with payment in rubles based on actual usage.