Vercel AI SDK Adds Support for xAI's Grok Imagine Video 1.5, Including 1080p Generation
Vercel released @ai-sdk/xai version 4.0.36, adding support for xAI's grok-imagine-video-1.5 model with native 1080p resolution and a new referenceVoiceIds parameter for reference-to-video audio. The update also fixes a routing bug that misdirected certain video generation requests.
Vercel shipped version 4.0.36 of the @ai-sdk/xai package, adding developer support for xAI's grok-imagine-video-1.5 model through the Vercel AI SDK. The patch, detailed in two commits on GitHub, expands video generation capabilities and fixes a reference-routing bug affecting the provider integration.
What changed
The update introduces the grok-imagine-video-1.5 model id to the SDK, along with native 1080p support for both text-to-video and image-to-video generation. Developers using the standard resolution: '1920x1080' parameter will now have that setting map directly to xAI's 1080p mode.
Reference-to-video generation — where a source image guides the output — remains capped at 720p. According to the changelog, if a developer requests 1080p in reference-to-video mode, the SDK automatically downgrades the request and issues a warning rather than failing outright.
The release also fixes a routing bug in the provider logic. Previously, any non-empty inputReferences array triggered reference-to-video mode, even if the array contained only non-image references. This caused the SDK to send an empty reference_images: [] payload with no usable reference data, silently breaking the request.
A second commit adds a new referenceVoiceIds parameter for reference-to-video generation. Developers can now pass up to three xAI preset voice ids (for example, ['eve']) and cite them in prompts using <AUDIO_0> through <AUDIO_2> tags. The SDK translates these into a reference_audios: [{ voice_id }] field sent to xAI's POST /v1/videos/generations endpoint, allowing generated video clips to incorporate specific preset voices.
Context
This is a client library update to Vercel's AI SDK, not a new model release from xAI. The grok-imagine-video-1.5 model itself is xAI's, and this patch simply brings the SDK's provider integration in line with the model's existing capabilities — notably higher-resolution output and voice-referenced audio. No pricing, benchmark, or context window details were included in the release notes, and Vercel's changelog does not disclose per-token or per-generation costs for the video endpoint.
What this means
For developers building on xAI's video generation stack through Vercel's SDK, this patch closes a functional gap: full 1080p output for direct text-to-video and image-to-video calls, plus a cleaner way to inject reference audio into generated clips. The bug fix around reference routing is arguably the more consequential change — silent failures where requests appeared to succeed but returned no usable video references are the kind of issue that erodes trust in an SDK integration. Teams already using @ai-sdk/xai for video workflows should upgrade to 4.0.36 to avoid the previous routing bug and to access the new resolution and voice-reference options.
Related Articles
ElevenLabs Launches Music v2.5, Adds API Access and Free Tier for AI-Generated Songs
ElevenLabs has released Music v2.5, an updated version of its ElevenMusic generator, now available through both the app and API. The company says blind testing with nearly 48,000 comparison pairs showed listeners preferred v2.5 over the prior version, particularly for R&B, Hip-Hop, and orchestral genres.
Perplexity Says It Runs End-to-End Engineering Systems on OpenAI's GPT-6 Astra
Perplexity says it has shifted core engineering workflows, including code changes and production monitoring, onto OpenAI's GPT-6 Astra model. The claim comes from an OpenAI-published case study with no independent benchmark data released.
Perplexity Deploys OpenAI's Astra Model for Autonomous Code and Systems Management
Perplexity is using an OpenAI model referred to as Astra to handle software changes, communications, and production monitoring with less frequent human check-ins. OpenAI published the case study; specific model specs and benchmarks have not been disclosed.
Augment Code Claims 4.5x Developer Output Increase From Internal 'Software Factory' of AI Agents
Augment Code says its internal 'software factory'—a network of specialized agents built on its Cosmos platform—drove a 4.5x increase in size-adjusted developer output and cut median PR merge time from 11.2 to 3.1 hours over nine months. The company frames this as evidence that once AI writes nearly all new code, the bottleneck shifts to review, verification, and incident response.
Comments
Loading...