AI Model Showdown: The Top Contenders for Text, Image & Video Generation

San Francisco, CA – Artificial intelligence is reshaping content creation, with new models emerging at an unprecedented pace. A recent study by Quora’s Poe provides insights into the most widely used AI models for text, image, and video generation, offering valuable data for professionals and enthusiasts alike.

With a flood of AI tools available, choosing the right model can be overwhelming. Poe’s study, which analyzed millions of interactions, identifies the leading AI models based on user adoption and trends. Whether you’re exploring AI for the first time or refining your toolset, understanding the landscape of text, image, and video generation can help optimize your workflow.

AI-powered text generation has long been dominated by OpenAI’s GPT models, but the competition is heating up. According to the study, Anthropic’s Claude models have gained significant traction, with usage now nearly equal to OpenAI’s models. The introduction of Claude 3.5 Sonnet has made Anthropic a major player, closing the gap between the two companies. Poe’s findings indicate that users frequently switch to newer models as they are released, even if they remain loyal to a particular provider. The transition from GPT-4 to GPT-4o and from Claude 3 to Claude 3.5 showcases this trend, with users consistently gravitating toward the latest, most advanced AI versions.

Beyond OpenAI and Anthropic, emerging players are making their mark. DeepSeek’s R1 and V3 models now account for 7% of all Poe text generation activity, signaling a growing interest in alternative AI solutions. Meanwhile, Google’s Gemini models have seen a modest decline in usage among Poe subscribers, yet remain a formidable presence in the industry.

AI-driven image generation has experienced a dramatic shift in recent months. Poe’s study reveals that DALL-E 3 and Stable Diffusion, once considered industry leaders, have seen their market share drop by 80%. The decline comes as more AI-powered image models enter the space, expanding user choices from just three major models to over 25 competing options. Among the newcomers, BlackForestLabs’ FLUX family has emerged as the dominant player, accounting for 40% of Poe’s AI image generation requests. Google’s Imagen3 models hold the second-largest share at 30%, underscoring their influence in the space. Smaller but innovative platforms such as Playground and Ideogram maintain a dedicated following by frequently updating their models, though their overall market share remains at 10%. As AI-generated art continues evolving, the demand for cutting-edge models with higher fidelity and creative flexibility is expected to shape the industry’s future.

Unlike text and image generation, AI-driven video creation is a relatively new but fast-growing field. Until late 2024, video generation was virtually nonexistent on Poe’s platform. However, it has since exploded in popularity, with at least eight different providers now offering AI video capabilities. Leading the way is Runway, whose video model accounts for 30–50% of all AI video requests. Despite maintaining a strong presence, Runway’s dominance is gradually fading as new players enter the market. One of the biggest disruptors is Google’s Veo-2, which has captured 40% of the market within a short span, reflecting how quickly user preferences evolve. Meanwhile, other competitors such as Kling-Pro v1.5, Hailuo-AI, HunyuanVideo, and Wan-2.1 collectively hold 15% of the market, signaling growing diversification in AI-generated video tools.

The Poe study highlights a consistent trend: users prioritize the newest models and are quick to abandon older tools if they don’t keep pace with advancements. This phenomenon is especially evident in text and image generation, where newer releases almost instantly replace their predecessors. For users seeking the best AI-generated text, images, or videos, the report suggests one takeaway: adopting the latest model is crucial. Older AI versions often become obsolete quickly as newer models offer improved accuracy, efficiency, and creative control.

However, cost could become a significant factor in future adoption rates. As AI models grow more sophisticated, they also tend to become more expensive, which could discourage some users from upgrading as frequently. If prices continue rising, the fast adoption cycle may slow down as users weigh the cost-benefit ratio of each new model. Poe has also announced plans to release future reports examining AI models based on specific tasks and pricing structures, helping users determine the best AI for their needs across various applications.

As AI technology evolves, the competition between leading models is intensifying. OpenAI and Anthropic remain locked in a tight race for text generation supremacy, while the image and video generation markets continue expanding with new players. For businesses, content creators, and AI enthusiasts, staying informed about which models are leading the industry is critical for leveraging the most effective AI tools. With innovation happening at an unprecedented pace, the landscape of AI content generation is poised for further transformation in the coming months. Whether you’re a writer, artist, marketer, or developer, understanding how AI models are evolving can help you make better choices about which tools to use for your creative and professional needs. As AI continues to reshape the way we create content, keeping up with emerging trends and shifting user preferences will be essential for staying ahead in this rapidly changing space.