Grok Voice 3.0 is now available with Speech to Speech capability at the same price as Grok Voice 2.0. The alias `grok-voice-latest` now automatically points to `grok-voice-3.0`, so existing sessions are upgraded automatically. The API is unchanged.

Grok 4.5 is now available on the xAI API, designed for coding, agentic tasks, and knowledge work. It features configurable reasoning effort levels (low, medium, or high, with high as default). It is priced at $2/1M input tokens and $6/1M output tokens.

Grok Bot has been launched as a durable AI teammate that works on a persistent cloud computer. It includes messaging, approvals, connectors, and routines capabilities, enabling it to function as a persistent autonomous agent.

Grok 4.7 is xAI's frontier model for coding, agentic tasks, and knowledge work, now available on the xAI API. It features a 500k context window, text and image inputs with text-only output, and no text output limit. It supports multiple reasoning effort levels (low, medium, high, xhigh) and always returns reasoning.encrypted_content on the Responses API.

Grok 4.7 is now available on the xAI API as a frontier model for coding, agentic tasks, and knowledge work. It features a 500k context window, text and image inputs with text-only output, and no text output limit. Reasoning effort supports low, medium, high (default), and xhigh levels, and it is also served on the US regional endpoint.

grok-voice-transcribe-2.0 is now available as an updated speech-to-text model. The previous version grok-voice-transcribe-1.0 remains available, and the default is still grok-voice-transcribe-1.0. Users can explicitly select either version via the Speech to Text API.

The grok-imagine-image-2.0 image API received several updates: the quality parameter now supports 'auto' as a value and is the new default (replacing 'medium'), image editing now accepts up to 5 source images per request (up from 3), and new aspect ratios 21:9 and 5:2 are now supported for both image generation and editing.

Grok 4.6 is xAI's frontier model focused on coding, agentic tasks, and knowledge work, now available on the xAI API. It features a 500k context window with text and image inputs and no text output limit. Reasoning effort levels include low, medium, high (default), and xhigh.

grok-imagine-video-1.5 now supports text-to-video, image-to-video, and reference-to-video modalities including optional preset voices. Native 1080p resolution is available for text-to-video and image-to-video. Text-to-video runs as text-to-image then image-to-video under the hood.

Grok Voice Think Fast 2.0 is now available with Speech to Speech capability. The grok-voice-latest alias will route to this model starting August 5, 2026.

Speech to Text now accepts a vad_threshold parameter available as a streaming query param and batch multipart field. This allows tuning of the voice-activity gate that skips non-speech audio, with lower values enabling transcription of quieter or noisier speech useful for narrowband telephony. Setting the value to 0 disables the gate entirely.

Grok 4.5 became available in the API console for EU users. This marks the regional expansion of the model to European Union users. See the Grok 4.5 overview for full details.

Grok Voice Think Fast 1.0 is now available via the Voice Agent API. This model is designed for fast voice interactions and can be accessed using the model identifier 'grok-voice-think-fast-1.0'.

xAI announced the Grok Imagine API, offering state-of-the-art video generation across quality, cost, and latency. This represents a new API capability for video generation built on the Grok platform.

Grok 4.1 is now available to all users on grok.com, X, and the iOS and Android apps. It is rolling out immediately in Auto mode and can be selected explicitly as 'Grok 4.1' in the model picker. A fast variant, Grok 4.1 Fast, was also introduced along with Agent Tools API for next-generation tool-calling agents.

Grok 4 is described as the most intelligent model in the world, featuring native tool use and real-time search integration. It is available to SuperGrok and Premium+ subscribers as well as through the xAI API. A new SuperGrok Heavy tier was also introduced with access to Grok 4 Heavy, the most powerful version of Grok 4.

Major capability upgrade. Image generation integrated. Beats Claude 3 Opus and GPT-4o on several benchmarks.

Significant reasoning improvements. Context window expanded to 128K tokens. Better at math and coding tasks.

xAI launches Grok. Real-time X data access. Designed to answer questions others refuse. Available to X Premium subscribers.