# Image AI Models for Game Studios > Compare 130 image generation models on Layer — quality, capabilities, and per-generation pricing side by side. Generate production-ready assets for your game. ## Image models (130) - [GPT Image 2.5 Sunburst](https://layer.ai/models/openai-gpt-image-2-5-sunburst) ([md](https://layer.ai/models/openai-gpt-image-2-5-sunburst.md)) by OpenAI: OpenAI's premium GPT Image 2.5 variant: extra precision across edits and 1024px to 4K output, trading longer generation times for tighter control. - [GPT Image 2.5 Flare](https://layer.ai/models/openai-gpt-image-2-5-flare) ([md](https://layer.ai/models/openai-gpt-image-2-5-flare.md)) by OpenAI: OpenAI's default GPT Image 2.5 variant: fast, high-quality generation with natural lighting, richer textures, and precise edits up to 4K. - [GPT Image 2](https://layer.ai/models/openai-gpt-image-2) ([md](https://layer.ai/models/openai-gpt-image-2.md)) by OpenAI: OpenAI's latest image model with stronger text rendering, UI generation, and photorealism. Native output up to 4K with three quality tiers. - [Nano Banana 2.1](https://layer.ai/models/google-nano-banana-2-1) ([md](https://layer.ai/models/google-nano-banana-2-1.md)) (aka Gemini Nano Banana 2.1) by Google: Google's Nano Banana 2.1 image model for text-to-image and prompt-guided edits. - [Grok Imagine Image 2.0](https://layer.ai/models/xai-grok-imagine-image-2.0) ([md](https://layer.ai/models/xai-grok-imagine-image-2.0.md)) by xAI: xAI's Grok Imagine Image 2.0 model for high-quality text-to-image generation and editing. - [Gemini 3.1 Flash Image](https://layer.ai/models/google-gemini-3-1-flash-image) ([md](https://layer.ai/models/google-gemini-3-1-flash-image.md)) (aka Nano Banana 2, Gemini 3.1 Image) by Google: Google's fast, high-quality image generation model with multimodal reasoning. - [Seedream 5.0 Pro](https://layer.ai/models/bytedance-seedream-5-pro) ([md](https://layer.ai/models/bytedance-seedream-5-pro.md)) by ByteDance: Top-tier reasoning-first image generation from ByteDance, with built-in online search. - [Muse Image 1.0](https://layer.ai/models/meta-muse-image) ([md](https://layer.ai/models/meta-muse-image.md)) by Meta: Meta's Muse Image model with faithful instruction-following and precise multi-reference editing. - [MAI Image 2.5](https://layer.ai/models/microsoft-mai-image-2-5) ([md](https://layer.ai/models/microsoft-mai-image-2-5.md)) by Microsoft: Microsoft's photorealistic image model with strong typography and native editing. - [GPT Image 1.5](https://layer.ai/models/openai-gpt-image-1-5) ([md](https://layer.ai/models/openai-gpt-image-1-5.md)) by OpenAI: GPT Image 1.5 is the latest image generation model from OpenAI, with better instruction following and adherence to prompts. - [MAI Image 2.5 Pro](https://layer.ai/models/microsoft-mai-image-2-5-pro) ([md](https://layer.ai/models/microsoft-mai-image-2-5-pro.md)) by Microsoft: Microsoft's flagship image model for high-fidelity generation and native editing. - [FLUX.3 Image](https://layer.ai/models/bfl-flux-3-image) ([md](https://layer.ai/models/bfl-flux-3-image.md)) by Black Forest Labs: Black Forest Labs' FLUX.3 image model for native 1K to 4K generation and multi-reference editing. - [Gemini 3 Pro Image](https://layer.ai/models/google-gemini-3-pro-image) ([md](https://layer.ai/models/google-gemini-3-pro-image.md)) (aka Nano Banana Pro) by Google: Google's state-of-the-art image generation and editing model. - [Gemini 3.1 Flash-Lite Image](https://layer.ai/models/google-gemini-3-1-flash-lite-image) ([md](https://layer.ai/models/google-gemini-3-1-flash-lite-image.md)) (aka Nano Banana Lite, Nano Banana 2 Lite) by Google: Google's fastest, most cost-efficient Gemini image generation model. - [Grok Imagine Image](https://layer.ai/models/xai-grok-imagine-image) ([md](https://layer.ai/models/xai-grok-imagine-image.md)) by xAI: xAI's image generation model capable of creating high-quality images from text prompts. - [Qwen Image 2.1](https://layer.ai/models/qwen-qwen-image-2-1) ([md](https://layer.ai/models/qwen-qwen-image-2-1.md)) by Qwen: Qwen Image 2.1 text-to-image and editing with configurable size, guidance, and seed. - [Luma UNI-1 Max](https://layer.ai/models/lumalabs-uni-1-max) ([md](https://layer.ai/models/lumalabs-uni-1-max.md)) by Luma AI: The highest-fidelity tier of Luma UNI-1, for hero-quality stills and premium edits. - [Ideogram V4.5](https://layer.ai/models/ideogram-v4.5) ([md](https://layer.ai/models/ideogram-v4.5.md)) by Ideogram: Ideogram's text-to-image and editing model with accurate typography and three quality tiers. - [Seedream 5.0 Lite](https://layer.ai/models/bytedance-seedream-5-lite) ([md](https://layer.ai/models/bytedance-seedream-5-lite.md)) by ByteDance: Fast, high-quality image generation from ByteDance, optimized for creative advertising. - [Qwen Image 2 Pro](https://layer.ai/models/qwen-qwen-image-2-pro) ([md](https://layer.ai/models/qwen-qwen-image-2-pro.md)) by Qwen: Qwen Image 2 Pro tier with higher quality text-to-image generation and enhanced detail. - [Wan 2.7 Pro](https://layer.ai/models/alibaba-wan-2-7-pro) ([md](https://layer.ai/models/alibaba-wan-2-7-pro.md)) by Alibaba: Alibaba WAN 2.7 Pro image model for high-quality text-to-image generation and multi-image editing. - [Luma UNI-1](https://layer.ai/models/lumalabs-uni-1) ([md](https://layer.ai/models/lumalabs-uni-1.md)) by Luma AI: Luma's unified image model for high-fidelity generation and prompt-based editing. - [Seedream 4.5](https://layer.ai/models/bytedance-seedream-4-5) ([md](https://layer.ai/models/bytedance-seedream-4-5.md)) by ByteDance: A new-generation image creation model from ByteDance, for both generation and editing. - [Krea 2 Large](https://layer.ai/models/krea-krea-2-large) ([md](https://layer.ai/models/krea-krea-2-large.md)) by Krea: Krea's flagship text-to-image model for high-fidelity generations with distinctive aesthetic range. - [FLUX.2 [flex]](https://layer.ai/models/bfl-flux-2-flex) ([md](https://layer.ai/models/bfl-flux-2-flex.md)) by Black Forest Labs - [Ideogram V4](https://layer.ai/models/ideogram-v4) ([md](https://layer.ai/models/ideogram-v4.md)) by Ideogram: Ideogram's latest text-to-image model with crisp visuals, accurate text, and image-to-image support. - [FLUX.2 [max]](https://layer.ai/models/bfl-flux-2-max) ([md](https://layer.ai/models/bfl-flux-2-max.md)) by Black Forest Labs - [FLUX.2 [dev] Edit](https://layer.ai/models/bfl-flux-2-dev-edit) ([md](https://layer.ai/models/bfl-flux-2-dev-edit.md)) by Black Forest Labs - [FLUX.2 [dev]](https://layer.ai/models/bfl-flux-2-dev) ([md](https://layer.ai/models/bfl-flux-2-dev.md)) by Black Forest Labs - [Krea 2 Turbo](https://layer.ai/models/krea-krea-2-turbo) ([md](https://layer.ai/models/krea-krea-2-turbo.md)) by Krea: Speed-optimized open-source version of Krea 2 — high-fidelity images in seconds, with the full Krea aesthetic range. - [FLUX.2 [klein] 9B](https://layer.ai/models/bfl-flux-2-klein-9b) ([md](https://layer.ai/models/bfl-flux-2-klein-9b.md)) by Black Forest Labs - [FLUX.2 [klein] 9B Edit](https://layer.ai/models/bfl-flux-2-klein-9b-edit) ([md](https://layer.ai/models/bfl-flux-2-klein-9b-edit.md)) by Black Forest Labs - [FLUX.2 [pro]](https://layer.ai/models/bfl-flux-2-pro) ([md](https://layer.ai/models/bfl-flux-2-pro.md)) by Black Forest Labs - [GPT Image 1](https://layer.ai/models/openai-gpt-image-1) ([md](https://layer.ai/models/openai-gpt-image-1.md)) by OpenAI: Powerful, versatile OpenAI image model for creative and professional apps. - [Qwen Image 2](https://layer.ai/models/qwen-qwen-image-2) ([md](https://layer.ai/models/qwen-qwen-image-2.md)) by Qwen: Qwen Image 2 standard text-to-image model with strong prompt adherence and diverse style support. - [Qwen-Image Edit 2511](https://layer.ai/models/qwen-qwen-image-edit-2511) ([md](https://layer.ai/models/qwen-qwen-image-edit-2511.md)) by Qwen: Latest Qwen image editing model. Supports prompt-guided transformations with enhanced quality. - [Qwen-Image Edit 2511 Multiple Angles](https://layer.ai/models/qwen-qwen-image-edit-2511-multiple-angles) ([md](https://layer.ai/models/qwen-qwen-image-edit-2511-multiple-angles.md)) by Qwen: Qwen Image Edit 2511 with the Multiple Angles LoRA. Re-renders the input image from a chosen camera angle. - [Gemini 2.5 Flash Image](https://layer.ai/models/google-gemini-2-5-flash-image) ([md](https://layer.ai/models/google-gemini-2-5-flash-image.md)) (aka Nano Banana) by Google: Google's image generation and editing model capable of multimodal reasoning. - [HiDream O1 Image 1.0](https://layer.ai/models/hidream-hidream-o1-image) ([md](https://layer.ai/models/hidream-hidream-o1-image.md)) by Hidream: HiDream's full O1 Image model — create, edit, and personalize images up to 2K in one native model. - [Qwen-Image Edit 2509](https://layer.ai/models/qwen-qwen-image-edit-2509) ([md](https://layer.ai/models/qwen-qwen-image-edit-2509.md)) by Qwen: Advanced image editing model. Enhanced performance for fine-grained manipulation. - [Kling Image V3](https://layer.ai/models/kling-image-v3) ([md](https://layer.ai/models/kling-image-v3.md)) by Kling: Kling V3 image model. High-quality images with negative prompts, supports up to 2K resolution. - [Z-Image Turbo](https://layer.ai/models/qwen-z-image-turbo) ([md](https://layer.ai/models/qwen-z-image-turbo.md)) by Qwen: Ultra-fast 6B parameter image model from Tongyi-MAI, optimized for near real-time generation. - [FLUX.2 [klein] 4B](https://layer.ai/models/bfl-flux-2-klein-4b) ([md](https://layer.ai/models/bfl-flux-2-klein-4b.md)) by Black Forest Labs - [FLUX.2 [klein] 4B Edit](https://layer.ai/models/bfl-flux-2-klein-4b-edit) ([md](https://layer.ai/models/bfl-flux-2-klein-4b-edit.md)) by Black Forest Labs - [FLUX.1 Kontext [max]](https://layer.ai/models/bfl-flux-1-kontext-max) ([md](https://layer.ai/models/bfl-flux-1-kontext-max.md)) by Black Forest Labs: Premium model for max editing performance, superior typography, and visual narrative consistency. - [FLUX.1 Kontext [pro]](https://layer.ai/models/bfl-flux-1-kontext-pro) ([md](https://layer.ai/models/bfl-flux-1-kontext-pro.md)) by Black Forest Labs: Pro-grade multimodal model for fast, iterative editing, style transfer, and consistency. - [Bria FIBO Edit Restore 1.0](https://layer.ai/models/bria-fibo-edit-restore-image) ([md](https://layer.ai/models/bria-fibo-edit-restore-image.md)) by Bria AI: Bria's one-shot photo restoration, trained on licensed data. - [Qwen-Image](https://layer.ai/models/qwen-qwen-image) ([md](https://layer.ai/models/qwen-qwen-image.md)) by Qwen: Open-source T2I model. Excels at high-res images from complex text, notable for clear, stylized text. - [FLUX.1 Kontext [dev]](https://layer.ai/models/bfl-flux-1-kontext-dev) ([md](https://layer.ai/models/bfl-flux-1-kontext-dev.md)) by Black Forest Labs: Open-weight, multimodal model for context-aware image editing. Excels at iterative edits via text. - [HiDream O1 Image Dev](https://layer.ai/models/hidream-hidream-o1-image-dev) ([md](https://layer.ai/models/hidream-hidream-o1-image-dev.md)) by Hidream: HiDream's distilled O1 Image Dev model — create, edit, and personalize images up to 2K in one native model. - [FLUX 1.1 [pro]](https://layer.ai/models/bfl-flux-1-1-pro) ([md](https://layer.ai/models/bfl-flux-1-1-pro.md)) by Black Forest Labs: Next-gen FLUX model. 6x faster with enhanced prompt adherence and top-tier quality for production. - [FLUX 1.1 [pro] Ultra](https://layer.ai/models/bfl-flux-1-1-pro-ultra) ([md](https://layer.ai/models/bfl-flux-1-1-pro-ultra.md)) by Black Forest Labs: Delivers ultra-high-res (up to 4MP) images with superior photorealism, detail, and speed. - [FLUX.1 Krea [dev]](https://layer.ai/models/bfl-flux-1-dev-krea) ([md](https://layer.ai/models/bfl-flux-1-dev-krea.md)) by Black Forest Labs: Open-weight model co-developed with Krea AI. Excels in photorealism and aesthetics. - [FLUX.1 SRPO [dev]](https://layer.ai/models/bfl-flux-1-dev-srpo) ([md](https://layer.ai/models/bfl-flux-1-dev-srpo.md)) by Black Forest Labs: 12B flow transformer fine-tuned with SRPO for exceptional photorealism and polished composition. - [FLUX.1 [dev]](https://layer.ai/models/bfl-flux-1-dev) ([md](https://layer.ai/models/bfl-flux-1-dev.md)) by Black Forest Labs: Powerful, open-weight 12B image model. Excels in image quality, prompt adherence, and commercial use. - [FLUX.1 [schnell]](https://layer.ai/models/bfl-flux-1-schnell) ([md](https://layer.ai/models/bfl-flux-1-schnell.md)) by Black Forest Labs: Ultra-fast, open-source model. Generates high-quality images in 1-4 steps for rapid prototyping. - [SeedVR2 Image Upscaler](https://layer.ai/models/bytedance-seedvr-upscale-image) ([md](https://layer.ai/models/bytedance-seedvr-upscale-image.md)) by ByteDance: ByteDance's SeedVR2 model for high-quality image upscaling. - [Qwen-Image 2512](https://layer.ai/models/qwen-qwen-image-2512) ([md](https://layer.ai/models/qwen-qwen-image-2512.md)) by Qwen: Latest Qwen text-to-image model with enhanced quality, detail, and prompt adherence. - [Topaz Enhance](https://layer.ai/models/topaz-enhance) ([md](https://layer.ai/models/topaz-enhance.md)) by Topaz Labs: Enhance image quality with advanced upscaling. - [Topaz Generative Enhance](https://layer.ai/models/topaz-enhance-generative) ([md](https://layer.ai/models/topaz-enhance-generative.md)) by Topaz Labs: Topaz Generative Enhance for enhanced details. - [ESRGAN Upscaler](https://layer.ai/models/esrgan-upscaler) ([md](https://layer.ai/models/esrgan-upscaler.md)) by ESRGAN: Very fast upscaling with good quality. - [Seedream 5.0 Pro Layerize](https://layer.ai/models/bytedance-seedream-5-pro-layerize) ([md](https://layer.ai/models/bytedance-seedream-5-pro-layerize.md)) by ByteDance: Split a finished image into independently editable layers with Seedream 5.0 Pro Layerize. - [Clarity Creative Upscaler](https://layer.ai/models/clarity-creative) ([md](https://layer.ai/models/clarity-creative.md)) by Clarity AI: Upscale images with high fidelity or creativity. - [Recraft Crisp Upscale](https://layer.ai/models/recraft-upscale-crisp) ([md](https://layer.ai/models/recraft-upscale-crisp.md)) by Recraft: Boost resolution while refining small details and faces. - [Topaz Standard V2](https://layer.ai/models/topaz-standard-v2-upscale-image) ([md](https://layer.ai/models/topaz-standard-v2-upscale-image.md)) by Topaz Labs: Topaz Gigapixel precision upscaling, faithful to the original image. - [Stable Diffusion 3](https://layer.ai/models/stabilityai-sd3) ([md](https://layer.ai/models/stabilityai-sd3.md)) (aka SD3) by Stability AI: Latest open-weight MM-DiT model. Major improvements in quality, prompt following, and text rendering. - [Recraft V3](https://layer.ai/models/recraft-v3) ([md](https://layer.ai/models/recraft-v3.md)) by Recraft: Versatile image model for graphic design. Generates legible, stylized text and scalable vector art (SVG). - [Ideogram Remove Background](https://layer.ai/models/ideogram-remove-background) ([md](https://layer.ai/models/ideogram-remove-background.md)) by Ideogram: Fast, clean background removal from Ideogram. - [Recraft Creative Upscale](https://layer.ai/models/recraft-upscale-creative) ([md](https://layer.ai/models/recraft-upscale-creative.md)) by Recraft: Upscale for a sharper, cleaner, higher-resolution result. - [Kling Image O3](https://layer.ai/models/kling-image-o3) ([md](https://layer.ai/models/kling-image-o3.md)) by Kling: Kling Omni 3 image model. High-quality images with text rendering capabilities up to 4K resolution. - [Krea 3](https://layer.ai/models/krea-krea-3) ([md](https://layer.ai/models/krea-krea-3.md)) by Krea: Krea 3 text-to-image and multi-reference editing at 1K or 2K. - [FLUX.3 Klein Base Edit](https://layer.ai/models/bfl-flux-3-klein-base-edit) ([md](https://layer.ai/models/bfl-flux-3-klein-base-edit.md)) by Black Forest Labs: Black Forest Labs' FLUX.3 Klein Base editor for multi-reference edits at 1K to 4K. - [FLUX.3 Klein Base](https://layer.ai/models/bfl-flux-3-klein-base) ([md](https://layer.ai/models/bfl-flux-3-klein-base.md)) by Black Forest Labs: Black Forest Labs' FLUX.3 Klein Base for native 1K to 4K text-to-image. - [FLUX.3 Klein](https://layer.ai/models/bfl-flux-3-klein) ([md](https://layer.ai/models/bfl-flux-3-klein.md)) by Black Forest Labs: Black Forest Labs' distilled FLUX.3 Klein model for fast 1K to 4K generation and multi-reference editing. - [Text Removal 1.0](https://layer.ai/models/image-editing-text-removal) ([md](https://layer.ai/models/image-editing-text-removal.md)) by fal: Remove all the text from an image. - [Bria FIBO Edit Rewrite Text 1.0](https://layer.ai/models/bria-fibo-edit-rewrite-text) ([md](https://layer.ai/models/bria-fibo-edit-rewrite-text.md)) by Bria AI: Rewrite the text in an image, keeping its typography. - [Bria Eraser 1.0](https://layer.ai/models/bria-erase-image) ([md](https://layer.ai/models/bria-erase-image.md)) by Bria AI: Bria's eraser for painted areas, trained on licensed data. - [Object Removal](https://layer.ai/models/fal-object-removal-erase-image) ([md](https://layer.ai/models/fal-object-removal-erase-image.md)) by fal: Low-cost object removal by name or by painted area. - [Ideogram Object Removal 3.0](https://layer.ai/models/ideogram-erase-image) ([md](https://layer.ai/models/ideogram-erase-image.md)) by Ideogram: Ideogram's prompt-free eraser for painted areas. - [Finegrain Eraser](https://layer.ai/models/finegrain-erase-image) ([md](https://layer.ai/models/finegrain-erase-image.md)) by Finegrain: Remove objects along with their shadows, reflections and lighting. - [Expression Edit 1.0](https://layer.ai/models/image-editing-expression-change) ([md](https://layer.ai/models/image-editing-expression-change.md)) by fal: Change a face's expression to the one you describe. - [Expression Change 2.0](https://layer.ai/models/image-apps-v2-expression-change) ([md](https://layer.ai/models/image-apps-v2-expression-change.md)) by fal: Give a face one of twenty named expressions. - [Image2Pixel](https://layer.ai/models/image2pixel) ([md](https://layer.ai/models/image2pixel.md)) by fal: Turn an image into pixel-perfect pixel art. - [Bria FIBO Edit Colorize 1.0](https://layer.ai/models/bria-fibo-edit-colorize-image) ([md](https://layer.ai/models/bria-fibo-edit-colorize-image.md)) by Bria AI: Bria's one-shot colorization, trained on licensed data. - [TeleStyle 2.0](https://layer.ai/models/telestyle-v2) ([md](https://layer.ai/models/telestyle-v2.md)) by Telestyle: Repaint an image in the style of a reference image. - [Ideogram Layerize Text 3.0](https://layer.ai/models/ideogram-v3-layerize-text) ([md](https://layer.ai/models/ideogram-v3-layerize-text.md)) by Ideogram: Split a flat graphic into a clean background and live text layers. - [PATINA Tileable 1.0](https://layer.ai/models/fal-patina-tileable) ([md](https://layer.ai/models/fal-patina-tileable.md)) by fal: Tileable PBR materials from a prompt or a photo: texture plus base color, normal, roughness, metalness and height. - [PATINA 1.0](https://layer.ai/models/fal-patina) ([md](https://layer.ai/models/fal-patina.md)) by fal: Read a texture's base color, normal, roughness, metalness and height maps, packed for any game engine. - [P-Image Upscale](https://layer.ai/models/pruna-p-image-upscale) ([md](https://layer.ai/models/pruna-p-image-upscale.md)) by Pruna AI: Pruna's sub-second upscaler, scaling images up to 8x and 128 megapixels. - [P-Image](https://layer.ai/models/pruna-p-image) ([md](https://layer.ai/models/pruna-p-image.md)) by Pruna AI: Pruna's sub-second text-to-image and multi-image editing model built for production volume. - [Seedream 5.0 Flash Layerize](https://layer.ai/models/bytedance-seedream-5-flash-layerize) ([md](https://layer.ai/models/bytedance-seedream-5-flash-layerize.md)) by ByteDance: Split a finished image into editable layers, fast, with Seedream 5.0 Flash Layerize. - [Seedream 5.0 Flash](https://layer.ai/models/bytedance-seedream-5-flash) ([md](https://layer.ai/models/bytedance-seedream-5-flash.md)) by ByteDance: Fast, budget-friendly image generation and editing from ByteDance's Seedream 5.0 line. - [Soul Cinema](https://layer.ai/models/higgsfield-soul-cinema) ([md](https://layer.ai/models/higgsfield-soul-cinema.md)) by Higgsfield: Higgsfield's text-to-image model with a fixed cinematic, film-still look. - [Soul 2](https://layer.ai/models/higgsfield-soul-2) ([md](https://layer.ai/models/higgsfield-soul-2.md)) by Higgsfield: Higgsfield's photographic text-to-image model for fashion, editorial and character imagery. - [Recraft V4 Pro Vector](https://layer.ai/models/recraft-v4-pro-vector) ([md](https://layer.ai/models/recraft-v4-pro-vector.md)) by Recraft: Recraft's pro V4 text-to-vector tier, with higher fidelity than the standard vector tier. - [Recraft V4 Vector](https://layer.ai/models/recraft-v4-vector) ([md](https://layer.ai/models/recraft-v4-vector.md)) by Recraft: Recraft's V4 text-to-vector model. Generates scalable vector art (SVG) directly from a prompt. - [Google Virtual Try-On 1.0](https://layer.ai/models/google-virtual-try-on) ([md](https://layer.ai/models/google-virtual-try-on.md)) by Google: Google Virtual Try-On: dress a person photo with a clothing product image. - [Topaz Dust-Scratch V2](https://layer.ai/models/topaz-dust-scratch-v2-restore-image) ([md](https://layer.ai/models/topaz-dust-scratch-v2-restore-image.md)) by Topaz Labs: Topaz film clean-up for dust and scratches. - [Topaz Recover 3 Restore](https://layer.ai/models/topaz-recover-3-restore-image) ([md](https://layer.ai/models/topaz-recover-3-restore-image.md)) by Topaz Labs: Topaz's generative restoration, rebuilding natural detail. - [Topaz Denoise Max](https://layer.ai/models/topaz-denoise-max-denoise-image) ([md](https://layer.ai/models/topaz-denoise-max-denoise-image.md)) by Topaz Labs: Topaz's generative denoiser, rebuilding detail as it cleans. - [Topaz Denoise Extreme](https://layer.ai/models/topaz-extreme-denoise-image) ([md](https://layer.ai/models/topaz-extreme-denoise-image.md)) by Topaz Labs: Topaz's most aggressive conventional denoiser. - [Topaz Denoise Strong](https://layer.ai/models/topaz-strong-denoise-image) ([md](https://layer.ai/models/topaz-strong-denoise-image.md)) by Topaz Labs: Topaz denoising tuned for heavier noise. - [Topaz Denoise Normal](https://layer.ai/models/topaz-normal-denoise-image) ([md](https://layer.ai/models/topaz-normal-denoise-image.md)) by Topaz Labs: Topaz general-purpose denoising at the source resolution. - [Topaz Recover 3](https://layer.ai/models/topaz-recover-3-upscale-image) ([md](https://layer.ai/models/topaz-recover-3-upscale-image.md)) by Topaz Labs: Topaz's restorative upscaler, rebuilding natural detail in damaged images. - [Topaz Recovery V2](https://layer.ai/models/topaz-recovery-v2-upscale-image) ([md](https://layer.ai/models/topaz-recovery-v2-upscale-image.md)) by Topaz Labs: Topaz's upscaler for extremely low-resolution sources. - [Topaz Transparency Upscale](https://layer.ai/models/topaz-transparent-upscale-image) ([md](https://layer.ai/models/topaz-transparent-upscale-image.md)) by Topaz Labs: Topaz upscaling that preserves the alpha channel end to end. - [Topaz Bloom Realism](https://layer.ai/models/topaz-bloom-realism-upscale-image) ([md](https://layer.ai/models/topaz-bloom-realism-upscale-image.md)) by Topaz Labs: Topaz's creative upscaler with its invented detail biased toward photorealism. - [Topaz Standard MAX](https://layer.ai/models/topaz-standard-max-upscale-image) ([md](https://layer.ai/models/topaz-standard-max-upscale-image.md)) by Topaz Labs: Topaz's precision-leaning generative upscaler. - [Topaz Wonder 3.5](https://layer.ai/models/topaz-wonder-3.5-upscale-image) ([md](https://layer.ai/models/topaz-wonder-3.5-upscale-image.md)) by Topaz Labs: Topaz's generative image upscaler, rebuilding detail with fewer repeated patterns. - [Topaz High Fidelity V3](https://layer.ai/models/topaz-high-fidelity-v3-upscale-image) ([md](https://layer.ai/models/topaz-high-fidelity-v3-upscale-image.md)) by Topaz Labs: Topaz's detail-preserving precision upscaler for professional photography. - [Topaz Bloom 2](https://layer.ai/models/topaz-bloom-2-upscale-image) ([md](https://layer.ai/models/topaz-bloom-2-upscale-image.md)) by Topaz Labs: Topaz's creative image upscaler, reinventing detail with an adjustable creativity dial. - [Topaz Low Resolution V2](https://layer.ai/models/topaz-low-resolution-v2-upscale-image) ([md](https://layer.ai/models/topaz-low-resolution-v2-upscale-image.md)) by Topaz Labs: Topaz's precision upscaler tuned for small, compressed sources. - [Topaz Text Refine](https://layer.ai/models/topaz-text-refine-upscale-image) ([md](https://layer.ai/models/topaz-text-refine-upscale-image.md)) by Topaz Labs: Topaz's precision upscaler that keeps text and hard shapes crisp. - [Topaz Redefine](https://layer.ai/models/topaz-redefine-upscale-image) ([md](https://layer.ai/models/topaz-redefine-upscale-image.md)) by Topaz Labs: Topaz's prompt-guided generative upscaler. - [Topaz CGI](https://layer.ai/models/topaz-cgi-upscale-image) ([md](https://layer.ai/models/topaz-cgi-upscale-image.md)) by Topaz Labs: Topaz's precision upscaler for rendered art and CG rather than photographs. - [FLUX Pro VTO](https://layer.ai/models/bfl-flux-pro-vto) ([md](https://layer.ai/models/bfl-flux-pro-vto.md)) by Black Forest Labs: Virtual try-on from Black Forest Labs: dress a person photo with a garment reference. - [SAM 3.1 Image](https://layer.ai/models/sam-3-1-image) ([md](https://layer.ai/models/sam-3-1-image.md)) by Meta: Meta's SAM 3.1 for fast multi-object image segmentation. - [Recraft V4.1 Vector](https://layer.ai/models/recraft-v4-1-vector) ([md](https://layer.ai/models/recraft-v4-1-vector.md)) by Recraft: Recraft's V4.1 text-to-vector model. Generates scalable vector art (SVG) directly from a prompt. - [Recraft V4.1 Pro Vector](https://layer.ai/models/recraft-v4-1-pro-vector) ([md](https://layer.ai/models/recraft-v4-1-pro-vector.md)) by Recraft: Recraft's V4.1 Pro text-to-vector model. Generates scalable vector art (SVG) directly from a prompt. - [Image to SVG](https://layer.ai/models/fal-image-to-svg) ([md](https://layer.ai/models/fal-image-to-svg.md)) by fal: Convert raster images to clean, scalable SVG vector graphics with fine-grained control over detail. - [Pixelcut Background Removal](https://layer.ai/models/pixelcut-background-removal) ([md](https://layer.ai/models/pixelcut-background-removal.md)) by Pixelcut: High-quality image background removal from Pixelcut. - [Recraft V4.1](https://layer.ai/models/recraft-v4-1) ([md](https://layer.ai/models/recraft-v4-1.md)) by Recraft: Professional design model with extended prompt support (10,000 characters) and advanced style customization. - [Recraft V4.1 Pro](https://layer.ai/models/recraft-v4-1-pro) ([md](https://layer.ai/models/recraft-v4-1-pro.md)) by Recraft: Professional design model with extended prompt support (10,000 characters) and advanced style customization. - [Recraft V4](https://layer.ai/models/recraft-v4) ([md](https://layer.ai/models/recraft-v4.md)) by Recraft: Professional design model with extended prompt support (10,000 characters) and advanced style customization. - [Recraft V4 Pro](https://layer.ai/models/recraft-v4-pro) ([md](https://layer.ai/models/recraft-v4-pro.md)) by Recraft: Recraft's pro V4 tier, with higher fidelity and finer detail than the standard tier. - [Qwen-Image Layered](https://layer.ai/models/qwen-qwen-image-layered) ([md](https://layer.ai/models/qwen-qwen-image-layered.md)) by Qwen: Split an image into layers using Qwen-Image Layered. - [Seedream 4.0](https://layer.ai/models/bytedance-seedream-4-0) ([md](https://layer.ai/models/bytedance-seedream-4-0.md)) by ByteDance: Unified architecture for image generation and editing. Allows fluid movement from concept to refinement. - [BiRefNet v2](https://layer.ai/models/nankai-birefnet-v2) ([md](https://layer.ai/models/nankai-birefnet-v2.md)) by Nankai University: High-quality image background removal. - [Recraft Vectorize Image](https://layer.ai/models/recraft-recraft-vectorize-image) ([md](https://layer.ai/models/recraft-recraft-vectorize-image.md)) by Recraft: Convert raster images to vector graphics using Recraft. - [FLUX.1 [pro]](https://layer.ai/models/bfl-flux-1-pro) ([md](https://layer.ai/models/bfl-flux-1-pro.md)) by Black Forest Labs: Flagship commercial T2I model offering superior prompt adherence, quality, and stylistic outputs. ## Frequently Asked Questions ### Which is the best AI image model? GPT Image 2.5 Sunburst by OpenAI leads Layer's image models with an Artificial Analysis Arena Elo of 1218, ahead of 55 other scored image models. ### What are the top image models on Layer? The top image models by Elo rating are: 1. GPT Image 2.5 Sunburst (Elo 1218), 2. GPT Image 2.5 Flare (Elo 1208), 3. GPT Image 2 (Elo 1186), 4. Nano Banana 2.1 (Elo 1172), 5. Grok Imagine Image 2.0 (Elo 1167). Ratings come from blind votes in the Artificial Analysis Arena. ### How are image models ranked on this page? Image models are ranked by their Artificial Analysis Arena Elo: people compare two generations made from the same prompt without being told which model made which, and a model preferred more often scores higher. An Elo is a standing inside one modality, so models are only ever ranked against others of their own kind. A model with no arena score yet sits below the scored ones, ordered by how much it is used on Layer and how recently it was released. ### How many image models does Layer have? Layer gives you 130 image generation models from 27 providers — Black Forest Labs, Topaz Labs and Recraft among them — in one workspace, on one balance, behind one API. ### Which image models can I train on my own artwork? 16 of them can be fine-tuned on a studio's own art, FLUX.2 [dev] Edit among them. Upload a reference set, train a style, and every generation from it carries your look rather than the model's default one. ### How much do image models cost? Nothing is charged per seat. Generating on Layer spends Creative Units, each model has its own rate, and you pay only for what you generate — every model page states its current price, and the pricing page lists them all. --- [View original](https://layer.ai/models/image) | [All models](https://layer.ai/models)