Voice STT Expands to 20 Languages
Claude CodeClaude Code v2.1.69 expands voice speech-to-text support from 10 to 20 languages, adding Russian, Polish, Turkish, Dutch, Ukrainian, Greek, Czech, Danish, Swedish, and Norwegian. Voice mode, rolled out gradually via the /voice command in early March 2026, now reaches a substantially broader global developer audience with this language expansion. The update also ships two stability fixes for the voice system, including a space bar stuck-state bug caused by a module loading race condition.
Sources & Mentions
5 external resources covering this update
Claude Code rolls out a voice mode capability
TechCrunch
Anthropic adding voice mode to Claude Code in gradual rollout
9to5Mac
Anthropic Rolls Out Voice Mode in Claude Code for Developers
MacObserver
Anthropic Rolls Out Voice Mode for Claude Code
Winbuzzer
Anthropic Release Notes — March 2026 Latest Updates
Releasebot
Claude Code Voice Mode Reaches 20 Languages
Claude Code v2.1.69, published March 4, 2026, doubles its voice speech-to-text language roster, bringing the total to 20 supported languages with 10 new additions.
The New Languages
The ten newly supported languages are: Russian, Polish, Turkish, Dutch, Ukrainian, Greek, Czech, Danish, Swedish, and Norwegian. These join the original ten from Claude Code's voice mode launch. The expansion covers major Eastern European, Scandinavian, and Central European developer communities, reflecting a deliberate push to make voice-driven coding accessible beyond English-speaking markets.
Voice Mode Context
Claude Code's /voice command was introduced in a gradual rollout that began reaching broader groups of users in early March 2026. The feature allows developers to hold the space bar to dictate prompts instead of typing, then release to submit — a workflow particularly useful during extended coding sessions or when working hands-free. Voice mode uses ElevenLabs technology for speech recognition and does not require an additional subscription beyond an existing Claude plan.
Stability Improvements
The v2.1.69 release also addresses two voice-specific bugs. A race condition where the space bar could get stuck after a failed voice activation — caused by a module loading race and a cold GrowthBook flag fetch — has been resolved. Additionally, the voice waveform cursor no longer covers the first letter of the suffix when dictating mid-input. Voice startup also benefits from the removal of a warmup pulse animation that was causing re-render stutter during initialization.