<h1 id="astronavi-v01-for-krea-2">AstroNavi V01 for Krea 2</h1>A cinematic species LoRA inspired by the Na’vi from the Avatar universe, created specifically for Krea 2.AstroNavi V01 is designed to reproduce the distinctive appearance of tall blue humanoid forest people while keeping the surrounding scene, clothing, era, lighting, pose, emotion, and environment flexible through prompting.The model can be used for cinematic jungle scenes, alternative or deleted movie moments, medieval fantasy settings, modern-day photography, emotional portraits, romantic scenes, action sequences, and completely original characters.<h2 id="activation-tag">Activation Tag</h2><code>AstroNaviV01K2T</code>Recommended natural prompt structure:<code>a female member of the AstroNaviV01K2T species</code><code>a male member of the AstroNaviV01K2T species</code><code>two members of the AstroNaviV01K2T species</code>The activation tag is intended to carry the core species characteristics, including the distinctive skin appearance, facial anatomy, eyes, ears, markings, and overall non-human morphology.For the best flexibility, describe the variable elements normally in the rest of the prompt:<ul><li>Gender and number of characters</li><li>Hairstyle and braid decorations</li><li>Clothing, jewelry, weapons, and body paint</li><li>Pose, expression, and interaction</li><li>Camera angle and framing</li><li>Environment and historical period</li><li>Individual light sources and color contrast</li><li>Cinematic, photographic, or fantasy presentation</li></ul><h2 id="tested-resolutions">Tested Resolutions</h2>The model has been successfully tested in both horizontal and vertical cinematic formats:<ul><li>16:9 at approximately 2 megapixels</li><li>everything in between </li><li>9:16 at approximately 2 megapixels</li></ul>Both formats generated reliably during testing without noticeable composition or anatomy problems caused by the higher resolution.The LoRA performs especially well with detailed Krea 2 prompts that define an asymmetrical composition, multiple realistic light sources, strong warm-and-cool color separation, atmospheric depth, material textures, camera perspective, and an emotional narrative moment.<h2 id="what-it-handles-well">What It Handles Well</h2><ul><li>Detailed cinematic close-ups</li><li>Medium shots and environmental portraits</li><li>Male and female characters</li><li>Couples and character interactions</li><li>Emotional facial expressions</li><li>Tribal and ceremonial clothing</li><li>Modern-day fashion and everyday environments</li><li>Medieval fantasy interpretations</li><li>Jungle action and exploration scenes</li><li>Romantic and sensual editorial photography</li><li>Bioluminescent environments</li><li>Firelight, moonlight, sunlight, practical lights, and mixed lighting</li><li>Horizontal and vertical cinematic compositions</li></ul><h2 id="prompting-recommendation">Prompting Recommendation</h2>Krea 2 responds best when the prompt describes the complete visual scene rather than relying only on short keywords.A strong prompt should include:Scene and emotional moment, camera distance and viewing angle, asymmetrical character placement, character appearance and styling, expression and gaze direction, pose and body language, clothing materials, foreground and background elements, primary light source, secondary fill or rim light, color contrast, atmosphere, lens character, depth of field, film grain, texture, and subtle environmental movement.Example structure:<code>A female member of the AstroNaviV01K2T species standing on the right outer third of a rain-soaked forest clearing, cinematic medium full shot from a low three-quarter angle, wearing layered woven garments, bead ornaments and long decorated braids, holding a carved bow with a focused expression, warm sunset rim light cutting through cool blue-green jungle mist, wet foliage reflecting light, deep environmental shadows, atmospheric particles, detailed skin and fabric texture, anamorphic lens character, shallow foreground focus and fine cinematic film grain.</code><h2 id="important-notes">Important Notes</h2>This is a fan-made creative model inspired by the visual language of the Avatar universe. It is intended for original artwork, experimentation, alternative scenes, personal characters, cinematic concepts, and transformative fan creations.The training dataset contains more female than male examples and is particularly strong in faces, eyes, skin appearance, expressions, braids, jewelry, and cinematic portrait framing. Male characters also work well, although detailed prompting helps produce greater variation.For broader facial diversity, describe age, hairstyle, expression, clothing, body proportions, role, and personality through the prompt while allowing the activation tag to control the shared species appearance.Have fun creating your own clans, warriors, hunters, healers, travelers, lovers, modern-day characters, medieval interpretations, and lost cinematic moments.Created and tested by Astroburner.

Files with SHA256: c7464034c58cac6b35258911b95708a596da8ddcd1987b00b95ec8a47a9914fe