Back to changelog

Subscription plans, batch processing & API key management

Subscription plans land in the dashboard this week alongside a proper batch processing UI, long-requested API key search and sort, and reasoning cost visibility in logs.

Batch processing

Batch is now a first-class feature in the dashboard. Filter models by batch capability, toggle a Batch only view, and inspect batch metadata — status, counts, cost, and timestamps — in a dedicated tab inside any log entry. Documentation covers the full workflow: submitting jobs, polling for completion, webhook setup, and quota guidance.

API key management

Search by name — Find the right key instantly without paging through long lists.
Sorting — Sort keys by usage, creation date, or expiration date, in either direction.
Pagination — Long key lists now paginate properly, including in the subscription key table.

Reasoning cost in logs

The Cost tab in log details now shows reasoning token costs separately when a model uses extended thinking — useful for tuning prompt design when using Claude or Gemini in reasoning mode.

Audio input in the playground

Attach and send audio files directly in the chat playground. Models that support audio input accept the file as a message attachment, play it back inline, and respond to the transcribed content. The model filter now includes an Audio Input capability option.

Back to changelog

Get Started

Cut your AI bill, not your usage

Route every request to the right model. Track every dollar you spend. Cut your LLM costs by up to 60%.