SpaceXAI launches Grok Voice Transcribe 2.0 for multilingual speech
SpaceXAI says Grok Voice Transcribe 2.0 improves multilingual speech-to-text, adds diarization and timestamps, and keeps pricing unchanged.
#
Key takeaways
- SpaceXAI announced Grok Voice Transcribe 2.0 on September 18, 2026.
- The model supports batch files and real-time streams.
- It detects languages automatically and can follow language switches in one recording.
- SpaceXAI says existing API integrations get the upgrade without code changes.
- Pricing stays at $0.10 per hour for batch transcription and $0.20 per hour for streaming.
SpaceXAI updates its speech-to-text model
SpaceXAI has announced Grok Voice Transcribe 2.0, its newest speech-to-text model. The company says it is built on the audio foundation model behind Grok Voice. It is designed for multilingual transcription and supports both batch files and real-time streams.
The release adds several transcription features. SpaceXAI says the model automatically detects languages, follows language switches during one recording, and includes diarization, timestamps, and key-term controls. The company also says existing speech-to-text API integrations receive the accuracy upgrade without code changes.
What SpaceXAI says improved
SpaceXAI reports that internal evaluations showed lower error rates across telephony, conversations, spoken credentials, and short multilingual commands. The company also gives a specific result for short phrases spanning 19 languages. It says word error rate fell from 20.6 percent in version 1.0 to 6.8 percent in version 2.0.
The source also says the model ranks first for accuracy among 32 streaming systems on the public Artificial Analysis leaderboard. It adds that some additional comparisons come from internal production sets. That means the performance picture includes both public and internal evaluation contexts.
Pricing and rollout
SpaceXAI says pricing remains unchanged. Batch transcription stays at $0.10 per hour, and streaming stays at $0.20 per hour. The company also plans to make version 2.0 the API default and deprecate version 1.0 in the coming weeks.
During the transition, customers can pin the earlier model. That gives teams a temporary option if they need to keep using version 1.0 while they adjust their workflows. The source does not provide any further rollout details.
Operational considerations
The update appears focused on compatibility and accuracy. SpaceXAI says current API integrations do not need code changes, which may reduce migration effort. The model's support for language switching, diarization, and timestamps may also help teams working with mixed-language audio.
The source also points to a transition period. Customers who rely on version 1.0 can keep it pinned for now, but SpaceXAI says it will be deprecated in the coming weeks. That makes version management an operational consideration for users of the API.
Morocco relevance
The source reports no Morocco-specific launch, pricing, partnership, or availability. For readers, the global lesson is simple: when a speech-to-text system supports multilingual audio and API continuity, migration planning matters.
Bottom line
Grok Voice Transcribe 2.0 is a multilingual speech-to-text update with broader language handling and new transcription controls. SpaceXAI says it improves accuracy, keeps pricing stable, and can be adopted without code changes for existing API users. The main near-term issue is the planned move from version 1.0 to version 2.0.
Follow us on Google
Add Intelligence Artificielle Maroc as a preferred source to see more of our relevant stories in Google Search.
What would you like to build?
We build custom AI platforms, SaaS products, intelligent business applications, and automation systems.
This form is for project inquiries, not general questions about artificial intelligence.
