Speech is Cheap Transcribe

**Judgment:** A strong, production-oriented ASR skill that wraps the Speech is Cheap API, with compelling pricing and feature depth for high-volume transcription pipelines. Best suited for users comfortable relying on an external SaaS vendor for speech-to-text. **What it does well** - Integrates cleanly into OpenClaw via a simple `asr.sh` CLI interface, taking URLs or local files and returning JSON. - Offers advanced transcription options (speakers, word timestamps, labels, streaming, language control, output formats, webhooks). - Designed for automation: suitable for agents, batch jobs, and continuous processing workflows. - Clear configuration model using `SIC_API_KEY` and a straightforward status-check command. **Key risks / limitations** - Depends entirely on an external proprietary service (speechischeap.com): outages, API changes, or pricing shifts will directly affect this skill. - All audio is sent to a third party; even with a `--private` flag, users must assess compliance, data residency, and confidentiality requirements. - Performance and accuracy claims (speed and relative cost vs other providers) are marketing statements and not independently validated here. - Requires users to manage and secure an API key; misconfiguration will cause failures, and overuse may incur unexpected costs. **Recommended scenarios** - High-volume or cost-sensitive ASR workloads where external SaaS usage is acceptable. - Agent-driven workflows needing reliable JSON transcription with diarization, timestamps, and flexible output formats. - Pipelines that benefit from webhooks and streaming for near-real-time or asynchronous processing. **Less ideal when** - You need on-premise or offline transcription with no external data transfer. - You handle highly sensitive audio and cannot use third-party ASR vendors without strict contractual and compliance guarantees. - You require a fully open-source, self-hosted transcription stack with no per-minute charges.