Voice Generation
ElevenLabs released CLI v1 on August 24, 2026. Here is what developers can actually build with it: batch TTS, agents-as-code, CI/CD pipelines, and machine-readable API calls.
Some links are partner links: if you subscribe through them, we may earn a commission, at no extra cost to you. The crowd verdicts stay independent.
ElevenLabs CLI v1 landed on August 24, 2026, giving developers direct terminal access to the same API that powers the platform's web studio. The tool turns voice generation into a scriptable, automatable workflow rather than a GUI-dependent process.
The bottom line on CLI v1
ElevenLabs CLI lets developers run text-to-speech, music generation, and conversational agent management entirely from the terminal. With 80% of GLAD-AI-TOR voters recommending ElevenLabs ($5/mo Starter), CLI v1 makes automation the next logical step.
What the CLI actually ships with #
The release includes three core command groups plus a developer-focused toolchain:
| Command | What it does | Typical use case |
|---|---|---|
elevenlabs text-to-speech convert |
Generates audio files from text input | Batch voiceover, CI/CD narration |
elevenlabs music compose |
Creates music tracks from text prompts | Automated jingle/background audio |
elevenlabs agents list |
Manages conversational AI agents | Fleet management, bulk migrations |
Installation differs by platform: brew install elevenlabs/tap/elevenlabs on macOS, a Scoop bucket on Windows, or a universal shell script via cURL for other systems. The CLI authenticates via API key, stored either as an environment variable or in a local config file.
The real payoff is the machine-readable interface. The --schema flag returns JSON contracts that describe required parameters and response shapes. Coding agents (think Cursor, Cline, or Claude Code) can parse these schemas and construct valid API calls without human intervention. Error messages follow the same pattern: JSON for machines, readable suggestions for humans.
Agents-as-code changes the deployment story #
ElevenLabs introduced "agents-as-code" as a first-class workflow. Developers pull agent configurations into local files, edit them alongside application code, preview changes with --dry-run, and push updates back to production.
For teams managing conversational AI across multiple customer organizations, this turns agent deployment into a git-based process. Version control, code review, and rollback all work the way they do for any other software artifact. A team running 50 voice agents no longer needs to click through 50 dashboard screens to update a shared greeting; one script handles the fleet.
The CLI also ships pre-built "skills" for common tasks: agent creation, speech generation, and call transcription. These are essentially workflow instructions that teach coding agents how to complete end-to-end jobs without manual prompting. A developer can ask their coding assistant to "create a new customer support agent with a British accent" and the CLI skill provides the scaffolding.
ElevenLabs
The realism benchmark for AI voices: best-in-class TTS and cloning at a premium price
Partner link. The crowd verdicts stay independent.
Batch processing and CI/CD integration #
The practical unlock for production teams is piping. A shell script can read a list of text files, pass each through elevenlabs text-to-speech convert, and write the resulting audio to a directory in one run. Regenerations no longer require clicking through the web UI, one file at a time.
Consider a documentation team maintaining 200 tutorial videos. When the product updates, script narration needs re-recording. With CLI v1, that workflow looks like:
for file in scripts/*.txt; do
elevenlabs text-to-speech convert --text-file "$file" --output "audio/$(basename "$file" .txt).mp3"
done
CI/CD pipelines can now include voice generation steps. A documentation repo could trigger audio narration on merge; a marketing team could generate localized voiceovers as part of their build process. GitHub Actions, GitLab CI, and Jenkins all integrate cleanly since the CLI runs anywhere Node or the standalone binary installs.
The tradeoff is cost visibility. ElevenLabs' credit system still applies, and the crowd notes that real costs can hit 2-3x the advertised per-character rate due to failed takes and regenerations. Automating generation makes it easier to burn through credits without noticing. Teams should build cost guards into their pipelines: credit balance checks, alerts on high-volume runs, and dry-run validation before production batches.
Voice quality remains the differentiator #
The CLI ships the same voice models available in the web studio. Eleven v3 still leads the market for realism: in blind listening tests, its output routinely passes as human speech. The Flash v2.5 model (32 languages, 32ms latency) handles real-time use cases like IVR and voice agents.
ElevenLabs offers 10,000+ voices in the community library. Instant voice cloning starts at the $5/mo Starter plan (upload a sample, get a clone). Professional Voice Cloning (30+ minutes of training audio, near-indistinguishable results) unlocks on Creator at $22/mo.
The platform's weak spots carry over to the CLI. Long narrations can drift accent or switch language mid-sentence; users report forcing regenerations until a clean take lands. The credit system punishes those retakes, and all unused credits wipe on subscription cancellation. Support is largely AI-bot driven.
How the CLI compares to web-only competitors #
| Tool | CLI/API | Starting price | Crowd score | Cloning |
|---|---|---|---|---|
| ElevenLabs | CLI v1 + full API | $5/mo | 80% recommend | From $5/mo |
| Murf AI | Falcon API (no CLI) | $19/mo | 67% recommend | Enterprise only |
| Speechify | API (no CLI) | $12/mo | 56% recommend | From $19/mo |
Murf's Falcon API offers $0.01/minute pay-as-you-go pricing with 55ms latency, but there is no command-line wrapper. The platform excels at e-learning narration with its built-in timeline editor, but voice cloning requires an Enterprise sales call. Teams needing self-serve cloning should look elsewhere.
Speechify's API exists for Studio, but the focus remains on the reading app (1,000+ voices, OCR scan-and-listen, accessibility features). Studio's credit system burns fast at scale, and the trial-to-paid conversion has drawn billing complaints.
ElevenLabs is the only major voice AI vendor shipping a dedicated CLI tool as of this writing. For developers already using the platform, CLI v1 removes friction. For teams evaluating vendors, it adds a deployment and automation story that competitors lack.
The full feature breakdown lives in the ElevenLabs vs Murf AI comparison.
Pricing and credit math #
ElevenLabs pricing starts at $5/mo (Starter, annual billing) or $6 month-to-month. Creator runs $22/mo with a 50% discount on the first month. The jump to Pro ($99/mo) is steep.
| Plan | Monthly cost | Credits | Rough audio time |
|---|---|---|---|
| Free | $0 | 10k/mo | ~10 min TTS |
| Starter | $5 | 30k/mo | ~30 min TTS |
| Creator | $22 | 100k/mo | ~100 min TTS |
| Pro | $99 | 500k/mo | ~500 min TTS |
Credit burn varies by voice model (v3 costs more than Flash) and by script length. Retakes eat credits. Teams budgeting for CLI automation should assume higher per-minute costs than the headline numbers suggest.
Rollover rules matter: unused credits carry for two billing cycles (up to 3x monthly allowance), then expire. Canceling a subscription wipes all credits immediately.
ElevenLabs
The realism benchmark for AI voices: best-in-class TTS and cloning at a premium price
Partner link. The crowd verdicts stay independent.
The verdict #
- CLI v1 is a production-grade tool, not a demo. Batch TTS, agent management, and machine-readable schemas ship on day one.
- Agents-as-code makes voice AI deployments git-native. Pull, edit, preview, push.
- Cost discipline matters more with automation. Credits burn faster when scripts run unattended. Build guards into pipelines.
- ElevenLabs remains the realism benchmark at $5/mo Starter, now with terminal-native workflows.
- Competitors offer APIs but not CLIs. Murf locks cloning behind Enterprise; Speechify focuses on reading, not developer automation.
For a full breakdown of features, pricing, and crowd votes, see the ElevenLabs tool page. Compare alternatives in the AI voice generator rankings or check out the best free AI voice generators guide.
Keep exploring
Every claim above is backed by the arena's live data: crowd votes, verified pricing, honest pros & cons.
More from the arena journal #
Voice Generation
Speechify SIMBA 3.2 Tops the Leaderboards: What Those Benchmarks Actually Measure
Speechify's SIMBA 3.2 claims #1 on Artificial Analysis. Here's what the benchmarks test, how they compare to real-world use, and whether the $6/M pricing holds up against ElevenLabs at $100/M.
Aug 26, 2026 · 4 min read
Voice Generation
ElevenLabs Credits Explained: Why Your Bill Runs Out Faster Than You Expect
A deep dive into ElevenLabs' credit system reveals why real costs can reach 2.8x the advertised rate. Learn how credits burn, what happens when you cancel, and how to budget accurately.
Jul 17, 2026 · 5 min read
Voice Generation
Murf vs LOVO: LOVO Has Shut Down (What to Use Instead)
LOVO AI shut down after its May 2026 Chapter 7 bankruptcy, so this matchup has one survivor. Murf ($19/mo) is the default for e-learning narration, and here is what to use if you wanted what LOVO offered.
Aug 17, 2026 · 4 min read