Big Pickle

Big Pickle sits inside OpenCode Zen's lineup as a proprietary API model positioned for developer productivity and code quality. It is presented as an AI-powered coding assistant whose design leans toward agentic workflows, with reasoning and tool calling highlighted as core strengths for breaking down multi-step programming tasks and invoking external functions. Structured output and configurable temperature further support deterministic generation, JSON-shaped responses, and fine-grained control over code sampling behavior, making the model suitable for editor integrations, scripted automation, and backend pipelines that rely on stable, parseable outputs.

The model is delivered through OpenCode Zen with a 200,000-token context window and a 32,000-token maximum output, giving it room to ingest sizable codebases or long issue threads and return substantial diffs or refactors in a single response. It is offered at zero cost per million tokens across input, output, and cache reads and writes, which removes a key budgeting constraint for high-volume coding workflows and continuous refactoring loops. The combination of agentic capabilities and an expansive context makes Big Pickle a practical fit for teams that want a free, reasoning-capable coding model accessible through a single proprietary API without managing their own inference infrastructure.