OpenAI: GPT-6.1 Sol (batch)

GPT-6.1 Sol is a text model from OpenAI, positioned as an upgrade to GPT-6 Sol within the GPT-6 series. It sits below the flagship GPT-6 Astra, so it is not the top of the line, but it inherits the family's focus on demanding professional work. The provider describes it as suited for agentic coding, computer use, and document-heavy professional tasks. If you are already working within OpenAI's GPT-6 generation, this is the mid-tier option rather than the most capable one.

The model accepts text, images, and files as input, and it returns text only. Its context window is 1,050,000 tokens, which is large enough to hold entire codebases, long document sets, or extended conversation histories in a single request. The maximum output is 128,000 tokens, so responses can be substantial rather than truncated to short answers. Because images and files are accepted alongside text, you can mix screenshots, scanned pages, or uploaded documents into the same prompt as written instructions.

Those capabilities point to a few practical patterns. Agentic coding is a natural fit: the model can read repository files, reason about changes, and call tools across a long working session. Computer use scenarios benefit from image input, since screenshots can be passed in directly. Document-heavy professional work, such as reviewing contracts, extracting structured data from reports, or summarizing large archives, is supported by the combination of a very large context window and file input. Structured outputs and tool calling make it usable as the reasoning core of an automated pipeline rather than only as a chat interface.

You call it through api14's OpenAI-compatible API using the model id "openai/gpt-6.1-sol:batch". The supported request parameters are include_reasoning, max_tokens, reasoning, reasoning_effort, response_format, seed, structured_outputs, tool_choice, tools, and verbosity. In practice that means you can control how much reasoning the model exposes or performs, cap output length with max_tokens, request JSON or schema-conforming responses through response_format and structured_outputs, define and steer function calls with tools and tool_choice, and make sampling more reproducible with seed. The verbosity parameter lets you tune how expansive the prose is.

One thing to weigh is that this is a batch variant, and it sits below GPT-6 Astra in the same series. If your workload needs the strongest reasoning OpenAI offers in this generation, Astra is the higher tier. If you need output modalities beyond text, this model does not provide them. For text-centric work where a large context window, file and image input, and tool use matter more than flagship-level reasoning, GPT-6.1 Sol is a reasonable default.