AI models FAQs

ContentPod uses a different model for each job. Writing runs on OpenAI's GPT-5.6 Terra, with GPT-5.6 Luna for quick, short pieces. AI Interview voice runs on GPT-Realtime 2.1 (with GPT-Realtime 2.1 Mini as a budget mode), transcription uses GPT Transcribe, and Two-Way Interview audio uses Gemini multi-speaker voices. For images you can choose from six models, and for video from six more. The full lineup is listed below.

Every writing tool, including Article Writer, Caption Writer, Email Composer, FAQ Builder, and Guide Writer, uses OpenAI's GPT-5.6 models. GPT-5.6 Terra handles long-form writing and transforms, and GPT-5.6 Luna handles short, fast pieces. All writing tools follow built-in humanizer rules that steer drafts away from common AI tells, while your brand voice still takes priority.

AI Image Studio offers six models on every plan: Nano Banana 2 (1 credit, fast), Nano Banana Pro (2 credits, up to 4K), GPT Image 2.5 Flare (2 credits, fast OpenAI images and edits), GPT Image 2.5 Sunburst (3 credits, premium quality, the best text rendering, and transparent backgrounds), Ideogram 4.0 (1 credit, accurate on-image text), and Recraft 4.1 (1 credit, graphics and logos in your brand colors). Ideogram 4.0 also powers lettering in Social Image Creator, Infographic Generator, and Logo Generator, and Recraft 4.1 is available in Logo Generator.

Video generation is included on Creator, Pro, and Team. The models are Veo 3 Fast (10 credits, 8 seconds), Veo 3.1 (15 credits, 8 seconds, best Google quality), Grok Video, built on xAI Grok Imagine Video 1.5 (12 credits, 1 to 15 seconds), Kling 3.0 Standard with sound (4, 7, or 10 credits for 5, 10, or 15 seconds), Kling 3.0 Pro for the sharpest motion (5, 9, or 14 credits), and Seedance 2.5 Cinematic for multi-shot, film-style scenes (23 or 46 credits for 5 or 10 seconds, Pro and Team only).

OpenAI shut down Sora 2 and Sora 2 Pro on September 24, 2026, so they are no longer available in ContentPod. For similar results, use Veo 3.1 for the highest-quality 8-second clips or Grok Video for clips up to 15 seconds. Kling 3.0 is a lower-cost option that includes sound, and Seedance 2.5 Cinematic covers multi-shot scenes on Pro and Team.

Start with the cheapest model that fits the job, then step up if you need more. For images, Nano Banana 2 is good for quick drafts, Ideogram 4.0 or GPT Image 2.5 Sunburst for anything with words on it, Recraft 4.1 for brand-color graphics, and Nano Banana Pro for high-resolution detail. For video, Kling 3.0 Standard is the lowest-cost way to test an idea, Veo 3.1 gives the best 8-second quality, and Seedance 2.5 Cinematic is for polished, multi-shot pieces.

Yes, at no extra credit cost. In AI Image Studio (Create mode) and Social Image Creator you can add up to 3 reference images on Nano Banana 2, Nano Banana Pro, and both GPT Image 2.5 models, and 1 on Ideogram 4.0. Recraft 4.1 does not take references. For video, Veo 3 Fast, Veo 3.1, and Seedance take up to 3 references (one becomes the starting frame), while Grok Video and Kling 3.0 take 1 as the starting frame. Upload JPEG, PNG, or WebP files up to 10 MB, or pick from your Content Library.

Multimodal models are designed to understand and connect multiple input types, such as text plus images, and sometimes audio. In practice, they can interpret a prompt, reference visuals, and context together to produce a more coherent output. Many products still chain multiple specialized models under the hood, but multimodal planning can improve alignment between what you describe and what gets generated.