HYPE OR FACT? · CHECKED 27 OCTOBER 2026
Gemini API CLI: Google's media models in one terminal command
OUR VERDICT
FactOur original verdict: “Fact vs hype: it's real, but beta.”
THE CLAIM
“Google's new Gemini API CLI lets the coding agent you already use (Claude Code, Codex, Cursor) generate images, video, speech and music, analyse files and run Deep Research, so coding agents can call Gemini's media models as a tool.”
What’s actually true
- It's an official, free, open-source (Apache 2.0) command-line tool, gemini-api, with modes built for coding agents (dry runs, usage reports, an agent mode).
- It's an experimental preview, and its own docs say it is 'not a supported Google product'.
- The '$1-3 per Deep Research run' figure going around isn't on Google's pricing page, so treat it as an estimate; Deep Research is billed by tokens.
Sources
Cost
The CLI is free; you pay Gemini API prices. Gemini 3.8 Flash and text-to-speech have a free tier; image generation is paid only (about $0.03-$0.11 per image), video about $0.10 per second, a Lyria song $0.08.
USE IT IF
People already using a coding agent who want voiceovers, thumbnails, subtitles or transcripts from the same workflow. Start with something free, such as transcribing a recording to SRT.
SKIP IT IF
Anyone who needs a supported, stable product, or who hasn't set a budget for image and video costs.
Related AI tools
Related
How we knowChecked on 27 October 2026. Verdicts: Fact = the claim holds up; Partly fact = real, but with caveats or overstated parts; Hype = the headline claim isn't supported. Products change quickly, so check the maker's site before you buy. Our methods & sources ↗