Anthropic: Claude Haiku 4.5 (batch)

Anthropic: Claude Haiku 4.5 (batch) is a text generation model from Anthropic, served through api14 under the model id anthropic/claude-haiku-4.5:batch. It belongs to the Claude Haiku line, which Anthropic describes as its fastest and most efficient family, offering near-frontier intelligence with lower latency than the larger Claude models. The model was released by the provider on 2025-10-15, and api14 exposes it through the same OpenAI-compatible interface as the rest of the catalogue.

The model accepts text, images, and files as input, and returns text only. It has a context window of 200,000 tokens and can produce up to 64,000 output tokens in a single response. The supported request parameters are include_reasoning, max_tokens, reasoning, response_format, stop, structured_outputs, temperature, tool_choice, tools, top_k, and top_p. That mix covers the usual sampling controls plus tool calling, structured output, and reasoning-related options, so you can shape both the format and the length of what comes back.

Those capabilities point to a few practical jobs. Large document sets, long conversations, and codebases fit comfortably inside the 200,000 token window, so summarisation, extraction, and question answering over lengthy material are reasonable fits. Because images and files are accepted as input, document and screenshot analysis can be handled in the same call as the text prompt. Tool calling and structured outputs make it usable as the decision layer in agent workflows or as a component that returns JSON for downstream services, and the 64,000 token output ceiling supports longer generated answers rather than short snippets.

Calling it through api14 works the same way as any other model on the platform: send a request to the chat completions endpoint, or use the OpenAI SDK with a base URL pointed at api14, and set the model field to anthropic/claude-haiku-4.5:batch. Include your api14 key in the Authorization header as a bearer token. The parameters listed above can be passed in the request body; tools and tool_choice enable function calling, response_format and structured_outputs constrain the output shape, and temperature, top_p, top_k, stop, and max_tokens control sampling and length.

The batch variant in the model id indicates batch processing rather than the standard synchronous path, so check the batch workflow on api14 before wiring it into a latency-sensitive loop. For workloads that need the strongest reasoning Anthropic offers, a larger Claude model may be a better fit, while Haiku 4.5 targets speed and efficiency. Output is text only, so anything requiring generated images or audio needs a different model.