SkytellsVideo modelMera 2.0 Avatar Pro
Mera 2.0 Avatar Pro is a specialized variant of our flagship Mera 2.0 AI video generation model, engineered to push the boundaries of photorealism, human expression, and cinematic video synthesis. Unlike conventional avatar models that rely on transcripts, Mera listens directly to the original audio itself. It captures tone, pitch, breathing, pauses, rhythm, and emotional nuances, translating them into naturally synchronized lip movements, facial expressions, eye movements, gestures, and body language. It understands not just what is said, but how it is said. Powered by Mera's next-generation architecture, it combines lifelike human generation, consistent subject identity, and emotionally coherent performances. Its motion generation captures physical properties such as weight, momentum, contact, and object interactions, creating scenes that feel grounded in the real world. Generate videos up to 3 minutes long with ultra-fast inference averaging approximately 30 seconds, bringing long-form, high-fidelity video creation closer to realtime. Evaluated against demanding multilingual challenges, including Arabic and Chinese, Mera pushes the boundaries of pronunciation, audio-driven expression, and visual fidelity—particularly in linguistically complex languages such as Modern Standard Arabic. With intelligent prompt enhancement and resolutions up to 4K, Avatar Pro brings expressive, photorealistic human performance to professional video creation. Mera 2.0 Avatar Pro — It doesn't just read the words. It hears the human behind them.
Model availability and access depend on your workspace and plan.
Mera 2.0 Avatar ProAbout the model
Mera 2.0 Avatar Pro, explained.
Mera 2.0 Avatar Pro is a specialized variant of our flagship Mera 2.0 AI video generation model, engineered to push the boundaries of photorealism, human expression, and cinematic video synthesis. Unlike conventional avatar models that rely on transcripts, Mera listens directly to the original audio itself. It captures tone, pitch, breathing, pauses, rhythm, and emotional nuances, translating them into naturally synchronized lip movements, facial expressions, eye movements, gestures, and body language. It understands not just what is said, but how it is said. Powered by Mera's next-generation architecture, it combines lifelike human generation, consistent subject identity, and emotionally coherent performances. Its motion generation captures physical properties such as weight, momentum, contact, and object interactions, creating scenes that feel grounded in the real world. Generate videos up to 3 minutes long with ultra-fast inference averaging approximately 30 seconds, bringing long-form, high-fidelity video creation closer to realtime. Evaluated against demanding multilingual challenges, including Arabic and Chinese, Mera pushes the boundaries of pronunciation, audio-driven expression, and visual fidelity—particularly in linguistically complex languages such as Modern Standard Arabic. With intelligent prompt enhancement and resolutions up to 4K, Avatar Pro brings expressive, photorealistic human performance to professional video creation. Mera 2.0 Avatar Pro — It doesn't just read the words. It hears the human behind them.
Supported capabilities
- Image to video
- Audio to video
- Text to speech
- Avatar
- Lipsync
What you get back. Your generated video appears directly in your Relam chat.
Working with Mera 2.0 Avatar Pro
Shape the result.
Seed
Optional inputRandom seed for reproducible generation.
Audio
Questions & answers
More about Mera 2.0 Avatar Pro.
Mera 2.0 Avatar Pro is a specialized variant of our flagship Mera 2.0 AI video generation model, engineered to push the boundaries of photorealism, human expression, and cinematic video synthesis.
Unlike conventional avatar models that rely on transcripts, Mera listens directly to the original audio itself. It captures tone, pitch, breathing, pauses, rhythm, and emotional nuances, translating them into naturally synchronized lip movements, facial expressions, eye movements, gestures, and body language. It understands not just what is said, but how it is said.
Powered by Mera's next-generation architecture, it combines lifelike human generation, consistent subject identity, and emotionally coherent performances. Its motion generation captures physical properties such as weight, momentum, contact, and object interactions, creating scenes that feel grounded in the real world.
Generate videos up to 3 minutes long with ultra-fast inference averaging approximately 30 seconds, bringing long-form, high-fidelity video creation closer to realtime.
Evaluated against demanding multilingual challenges, including Arabic and Chinese, Mera pushes the boundaries of pronunciation, audio-driven expression, and visual fidelity—particularly in linguistically complex languages such as Modern Standard Arabic.
With intelligent prompt enhancement and resolutions up to 4K, Avatar Pro brings expressive, photorealistic human performance to professional video creation.
Mera 2.0 Avatar Pro — It doesn't just read the words. It hears the human behind them.
Skytells is the vendor behind Mera 2.0 Avatar Pro, a video model available in the Relam catalogue.
Random seed for reproducible generation.
Speech audio, up to 3 minutes; the video matches its length. Overrides voice_script.
Portrait of the person who speaks (first frame). Supports jpg, jpeg, png, webp. This input is required.
What the video should show.
Skytells adaptive inference setting for lip-sync under extreme motion.
Delivery instructions for the TTS voice (tone, pacing, accent, emotion). Not spoken.
Exact words to speak with Gemini TTS. Required when no audio is uploaded.
Let a vision model look at the image and append the scene motion to your prompt. Your words stay unchanged and come first.
Block content in the toggleable moderation categories. The always-on rules apply either way.
Gemini TTS voice for voice_script. Supported values are Zephyr (Female), Puck (Male), Charon (Male), Kore (Female), Fenrir (Male), Leda (Female), Orus (Male), Aoede (Female), Callirrhoe (Female), Autonoe (Female), Enceladus (Male), Iapetus (Male), Umbriel (Male), Algenib (Male), Despina (Female), Erinome (Female), Laomedeia (Female), Achernar (Female), Algieba (Male), Schedar (Male), Gacrux (Female), Pulcherrima (Female), Achird (Male), Zubenelgenubi (Male), Vindemiatrix (Female), Sadachbia (Male), Sadaltager (Male), Sulafat (Female), Alnilam (Male), Rasalgethi (Male). The model default is Zephyr (Female). Your workspace may offer a narrower selection depending on your plan.
Video resolution. 1080p, 2k and 4k generate at 768p and upscale it (24 fps kept); 4k is limited to 125 s of audio. Supported values are 480p, 768p, 1080p, 2k, 4k. The model default is 768p. Your workspace may offer a narrower selection depending on your plan.
Language of voice_script. Supported values are English (US), English (UK), Spanish, French, German, Italian, Portuguese (Brazil), Japanese, Korean, Hindi. The model default is English (US). Your workspace may offer a narrower selection depending on your plan.
Open your Relam workspace and look for Mera 2.0 Avatar Pro in the video model selector. Its required inputs are image. Available models, controls and access depend on your workspace and plan.