Zelili AI

Grok Imagine Tops Artificial Analysis AI Video Generation Leaderboards In Text-to-Video and Image-to-Video

Grok Imagine Tops Artificial Analysis AI Video Generation Leaderboards

Summary Box Summary Box [In a hurry? Just read this⚡]

  • xAI‘s Grok Imagine Video has taken the #1 spot on both Text-to-Video and Image-to-Video quality leaderboards from Artificial Analysis.
  • It achieved the highest ELO score in blind user preference tests: 1,245 for text-to-video and 1,336 for image-to-video, outperforming models from Runway, KlingAI, Google Veo, and OpenAI Sora.
  • Competitive API pricing at $4.20 per minute makes it more affordable than many premium alternatives like Sora 2 Pro ($30/min).
  • The model excels in visual fidelity, motion smoothness, and prompt adherence, launched in January 2026 with a free trial available via the Grok app.
  • Rankings are based on thousands of real user votes, showing xAI is now a serious contender in generative video AI alongside established players.

Grok Imagine ranking in video generation: xAI has achieved a notable milestone with its Grok Imagine Video model, securing the top spot on both the Text-to-Video and Image-to-Video quality leaderboards from Artificial Analysis.

This independent benchmarking platform evaluates AI video models through blind user preferences, using ELO scores to rank performance in generating coherent, high-quality videos from text prompts or input images.

Higher ELO scores indicate stronger overall quality, with evaluations based on thousands of samples.

Released in January 2026, Grok Imagine Video stands out for its balance of creativity, realism, and efficiency, making it accessible via the Grok app with a free trial option.

The rankings highlight the rapid advancements in generative video technology, where models compete on visual fidelity, motion smoothness, and adherence to prompts.

Text-to-Video Leaderboard Highlights

In the Text-to-Video category, models are assessed on their ability to create videos directly from descriptive text. Here’s the top 10 based on current ELO rankings:

RankCompanyModelELO Score95% CISamplesRelease DateAPI Pricing ($/min)
1xAIgrok-imagine-video1,245-12/123,566Jan 20264.20
2ViduVidu Q3 Pro1,242-12/123,386Jan 20269.60
3RunwayRunway Gen-4.51,236-7/79,681Dec 2025No API
4KlingAIKling 2.5 Turbo 1080p1,227-9/95,968Sept 20254.20
5GoogleVeo 3.1 Fast Preview1,226-9/95,914Oct 20256.00
6GoogleVeo 31,225-11/113,815Jul 202512.00
7GoogleVeo 3.1 Preview1,223-9/95,800Oct 202512.00
8KlingAIKling 2.6 Pro (January)1,216-13/132,817Jan 20264.20
9KlingAIKling O1 Pro (January)1,211-13/132,919Jan 202610.08
10OpenAISora 2 Pro1,210-9/96,141Sept 202530.00

Image-to-Video Leaderboard Highlights

For Image-to-Video, models animate static images into dynamic videos. The top 10 include:

RankCompanyModelELO Score95% CISamplesRelease DateAPI Pricing ($/min)
1xAIgrok-imagine-video1,336-12/123,569Jan 20264.20
2KlingAIKling 2.5 Turbo 1080p1,305-9/95,769Sept 20254.20
3GoogleVeo 3.1 Fast Preview1,301-9/95,953Oct 20256.00
4ViduVidu Q3 Pro1,298-12/123,414Jan 20269.60
5GoogleVeo 3.1 Preview1,297-9/95,953Oct 202512.00
6TeleAITeleVideo 2.01,278-12/123,069Jan 2026No API
7PixVersePixVerse V5.51,278-11/113,884Dec 20256.40
8RunwayRunway Gen-4.51,275-13/132,995Nov 2025No API
9PixVersePixVerse V51,275-8/86,363Aug 20256.40
10KlingAIKling 2.6 Pro (January)1,273-12/123,392Jan 20264.20

Implications for Users and the Industry

These rankings provide practical guidance for creators, marketers, and developers selecting video generation tools. Key considerations include:

  • Pricing Accessibility: Grok Imagine Video offers competitive API pricing at $4.20 per minute, undercutting premium options like OpenAI’s Sora 2 Pro at $30 per minute.
  • Speed vs. Quality Trade-offs: Models like Google’s Veo 3.1 Fast Preview prioritize quicker generation, ideal for rapid prototyping.
  • Recent Innovations: Many top entries, such as Vidu Q3 Pro and Kling 2.6 variants, were released in January 2026, reflecting intense competition.
  • Evaluation Reliability: ELO scores derive from crowdsourced blind votes, ensuring unbiased assessments, though confidence intervals show some variability.
  • Future Trends: Dominance in both categories suggests xAI is pushing boundaries in multimodal AI, potentially influencing applications in advertising, education, and entertainment.

Read More: Is Anthropic’s Claude Opus 4.6 on the Way? Here’s What We Know So Far