> ## Documentation Index > Fetch the complete documentation index at: https://documentation.api.odyssey.ml/llms.txt > Use this file to discover all available pages before exploring further. # Interaction Tips > A practical guide with examples to get you started interacting with Odyssey-2 Pro. ## Let's structure a prompt Consider including the below elements in your prompt. The more exhaustive your prompt, the clearer the picture Odyssey-2 Pro will be able to reference. 1. **Subject**: Objects, animals, people, or sceneries to be included. For example, *cityscape*, *desert*, *boats*, *puppies*, or *3 friends*. 2. **Action**: What the subject(s) are doing. For example, *walking*, *running*, *talking*, *painting*, or *sitting*. 3. **Environment:** The background, setting or time of day. 4. **Style**: Define art styles, artists, art mediums, colloquial aesthetics, or film style keywords; *horror*, *film noir*, or *cartoon*. 5. **Camera position and movement**: *top-down*, *aerial*, *eye-level*, *worm’s-eye*, etc. 6. **Composition**: How the shot is framed, such as *wide shot*, *close-up*, *single-shot* or *two-shot*. 7. **Focus and lens effects**: Use terms like *shallow focus*, *deep focus*, *soft focus*, *macro lens*, and *wide-angle lens* to achieve specific visual effects. 8. **Ambiance/mood/lighting:** Lighting and its color such as *soft light*, *harsh*, *neon*, *sunset*, *dark*, *ominous*, *blue tones*, *warm tones*, *day* or *night*. For example: > [Two Italian brothers dressed in steampunk, sitting in a booth at a harshly lit, futuristic, cyberpunk coffee cafe.](https://experience.odyssey.ml/CBVetKeYHh) *** ## Adding styles & visual aesthetics Odyssey-2 Pro understands a wide range of **colloquial stylistic descriptions**, even when they aren’t formal art terms. You can reference **your favorite artist**, **a well-known art movement**, or even **cultural shorthand** like “Minecraft style,” “1960s cartoon,” or “GTA 6 graphics.” These light-touch style cues are often enough for Odyssey-2 Pro to infer the entire aesthetic: color palettes, brushwork, rendering modes, lighting, and even cultural conventions. The examples below demonstrate some of the categories Odyssey-2 Pro responds to. ### [Pop art](https://experience.odyssey.ml/V3cbvheT27) Bright colors, bold outlines, graphic/comic vibes. Great for stylized, attention-grabbing scenes. ### [Surrealism](https://experience.odyssey.ml/wsZfL3SNJO) Dreamlike, bizarre, subconscious imagery that feels symbolic or dream-logic driven. ### [Cubism](https://experience.odyssey.ml/7CzLiATpWN) Fragmented geometry, multiple viewpoints at once. Odyssey-2 Pro leans into angular shapes and abstract composition. ### [Graffiti, street art](https://experience.odyssey.ml/n7pQ6qopFO) Sprayed textures, bold stencils, urban flair. Ideal for edgy or contemporary aesthetics. ### Other recognized styles * **Minimalism:** Clean shapes, limited palette, visual simplicity. Useful when you want clarity and negative space. * **Renaissance style:** Classical realism, dramatic lighting, and idealized forms; Odyssey-2 Pro applies period-accurate rendering. * **Expressionism:** Emotion-driven distortion and heavy brushstrokes. Odyssey-2 Pro picks up on intensity and exaggeration. Odyssey-2 Pro is flexible with colloquial or genre-based styles, even when they aren’t classical art movements. These examples show that Odyssey-2 Pro understands **cultural styles** just as well as formal art terms: * [**Photorealistic**](https://experience.odyssey.ml/QMnVCCvi3Y) * [**8-bit pixel art**](https://experience.odyssey.ml/yqYoHlRl7f) * [**Minecraft voxel style**](https://experience.odyssey.ml/z14NkDFT6D) * [**Japanese anime**](https://experience.odyssey.ml/UcWmD9kHBk) * [**1960s cartoon**](https://experience.odyssey.ml/wXBARO5s6K) * [**2010s cartoon**](https://experience.odyssey.ml/Nal9zxgTt8) * [**GTA-6 realism**](https://experience.odyssey.ml/xTbYtpTt10) ### Mixing multiple styles & visual concepts in one prompt Odyssey-2 Pro can faithfully blend multiple styles and concepts into one creation—often with one simple stylistic direction. For example, this single base prompt: > A man and his dog walking through an alley way towards the camera in soft focus, in the style… > > * [**dark fantasy**](https://experience.odyssey.ml/MIGedqjAKY) > * [**film noir**](https://experience.odyssey.ml/5Fw5ST3jRF) Notice how the stylistic cue alone affects the setting, the man, his clothes, the overall color palette, and the dog. Try prompting a fusion of your favorite representations. *** ## Framing your subject Odyssey-2 Pro understands common cinematography terms for **subject distance** and **framing**. These influence how much of the character fills the frame and are great for controlling intimacy, tension, or detail. * **Macro shot:** extreme detail of tiny subjects * **Extreme close-up (ECU):** detailed level * **Close-up (CU):** head fills frame * **Medium close-up (MCU):** chest-up * **Over-the-shoulder (OTS):** positioned behind one subject, looking over their shoulder toward another subject or focal point *** ## Adjusting the camera's position For storytelling or blocking, you can specify where the camera physically sits. This helps you choreograph generations with more precision. * **Bird’s-eye view:** Overhead, looking down upon subject * **Profile shot:** Side view of subject * **Back/behind shot:** Rear view of subject *** ## Using negative prompts Prompting Odyssey-2 Pro what you **don’t** want can help constrain and exclude attributes from generation. For example: | Prompt | Result | | ------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------ | | Animation of a large, solitary oak tree with leaves blowing vigorously in a strong wind. | [Standing Strong in the Storm](https://experience.odyssey.ml/mGDiJImnmk) | | Animation of a large, solitary oak tree with leaves blowing vigorously in a strong wind. Negative prompt: dark, stormy, or threatening atmosphere | [Gentle Wind, Silent Oak](https://experience.odyssey.ml/6PcdjfD6kQ) | *** # Tips for prompting midstream Dynamic present-tense action verbs (“puts on glasses”) can cause actions to loop, while **stative present-continuous descriptions** (“is wearing glasses”) describe a **completed, ongoing state**—a condition that is already true, not something that needs to repeat. **Action phrasing (verbs) → loops** **State description → no loops** *** # Known limitations Known non-real subjects and actions limit Odyssey-2 Pro’s ability to adhere to stylistic requests and realism. The model will often seek to accurately represent your subject/action, at the expense of your stylistic requests. Below, notice as we get farther from **“man”**, the silhouette becomes less a silhouette and realism suffers. | Prompt | Result | | ---------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------ | | Silhouette of **`a man`** standing in front of a bright setting sun side angle, cinematic contrast. | [Silhouette at Sunset Cliffs](https://experience.odyssey.ml/EMUJx6Tlzx) | | Silhouette of **`an Italian plumber`** standing in front of a bright setting sun side angle, cinematic contrast. | [Silhouette of a Hero at Dusk](https://experience.odyssey.ml/ody7nrOfvl) | | Silhouette of **`Mario`** standing in front of a bright setting sun side angle, cinematic contrast. | [Mario at Sunset's Edge](https://experience.odyssey.ml/Mnq2Z4DhOb) |