Explore Models

Original Aion model, robust for general tasks.

Latest model from AionLabs with improved reasoning.

A relaxed, informal male voice for chatting with friends.

A noble, chivalrous young male voice for heroic tales.

Remove image backgrounds with AI precision.

A cheerful, bubbly young female voice that radiates positivity.

Balanced speed and quality for daily conversations.

Improved Haiku with better reasoning.

Always points to the latest Anthropic Claude Haiku model (currently Claude Haiku 4.5).

Fast and lightweight model for quick responses.

Highly capable model for complex reasoning and creative tasks.

Exceptional performance on complex tasks, with enhanced reasoning and analysis.

Stable version with consistent performance for various applications.

Creative writing and deep analysis capabilities with long context window.

High-speed variant of Opus 4.6, optimized for faster responses with identical capabilities. 1M context window.

Anthropic's most capable model with high-resolution vision (up to 3.75MP), stronger coding, and enhanced safeguards.

Always points to the latest Anthropic Claude Sonnet model (currently Claude Sonnet 4.6).

Older but still effective version for daily tasks.

Fast and efficient model, with 10 free messages every 6 hours.

A deep, authoritative male voice ideal for professional presentations.

Specialized chat model from DeepSeek, optimized for conversation.

Text generation model from DeepSeek, suitable for general tasks.

Distilled reasoning model based on Llama 70B.

Distilled reasoning model based on Qwen 32B.

Special version optimized for speed.

DeepSeek V4 Flash. Fast, cheap, still near-frontier. Flat rate.

DeepSeek V4 Pro. 1M context, SOTA reasoning. Flat rate.

Colorize black and white images.

A serene, calm female voice in Indonesian.

An energetic, uplifting young female voice full of motivation.

Baidu's text-to-image model. Supports English, Chinese, and Japanese.

Fast 8-step distilled ERNIE image generation.

A polite, well-mannered young male voice for formal settings.

Professional Flux model with excellent prompt adherence.

Advanced image editing with improved prompt understanding.

Professional image editing with aspect ratio control.

12B parameter model, supports img2img with prompt strength.

Fastest Flux endpoint, optimized by PrunaAI.

Professional inpainting to remove objects.

Premium text-based image editing with max performance.

Text-based image editing with natural language.

Fast Flux model for quick iterations.

Flexible model for various styles.

Fast and economical, good for drafts.

Maximum capability Flux model.

Top-tier Flux model for demanding tasks.

Google's fast image model.

Google's multimodal model with deep understanding and search integration.

Google's latest high-quality image model preview.

Latest Gemini image model preview.

Fast model for immediate responses, suitable for light chat.

Latest preview of Gemini model with speed improvements.

Always points to the latest Google Gemini Flash model (currently Gemini 3.1 Flash).

Always points to the latest Google Gemini Pro model (currently Gemini 3.1 Pro).

Restore old, damaged, or low-quality faces with AI.

GLM 5.1 specialized for reasoning.

Z.AI's first multimodal coding model, natively handles image, video, and text inputs for vision-based coding and agent-driven tasks.

Always points to the latest OpenAI GPT model (currently GPT-5.5).

OpenAI's latest Instant chat model used in ChatGPT. Always resolves to the newest version.

OpenAI GPT Image 2 text-to-image. High quality, up to 4K. Token-based pricing.

OpenAI GPT Image 2 editing. Modify images with text prompts. Supports multiple reference images.

Always points to the latest OpenAI GPT Mini model.

Multimodal model with vision and audio capabilities.

Compact version of GPT-4o with fast responses and high quality.

Compact model with good performance for everyday use.

Ultra-lightweight model for simple tasks.

Economical model for lightweight tasks, suitable for prototyping.

Specialized for code generation and programming tasks.

Fifth-generation GPT, fast responses with high quality.

Advanced language model with high accuracy and deep context understanding.

Latest GPT-5.5 standard. 1M context, 128K max output. Charged per token.

OpenAI's most powerful model. 1M context, 128K max output. Charged per token.

A gentle, soothing female voice ideal for relaxation and meditation.

xAI's first image model based on Aurora architecture. Exceptional photorealism.

Optimized for speed, ideal for quick responses.

Agentic tool calling model with 2M context window and configurable reasoning.

Latest Grok model with enhanced capabilities.

Multi-agent model from xAI with advanced reasoning and collaboration.

xAI's Grok 4.3 with 1M context, built-in reasoning, and agentic capabilities. 40% cheaper input than Grok 4.20.

High-realism image generation from xAI.

xAI's flagship text-to-image model with exceptional aesthetic quality.

xAI's image editing model. Precise control over edits with text instructions.

xAI's flagship text-to-video model. Generates cinematic clips up to 15s with native synchronized audio and realistic motion.

Edit existing videos with text instructions. Output keeps original duration.

Extend a video by 6 or 10 seconds with smooth continuity.

Animate static images into short video clips with xAI's video model.

Generate short video clips from text descriptions.

High quality image-to-video with fine details.

Faster version of Hailuo 2.3, good for quick iterations.

Edit video with text instructions. Optional reference images (0-9).

Extend a video with Alibaba's HappyHorse. Up to 15s total.

Animate an image with optional prompt guidance. 3-15s.

Generate video from 1-9 reference images and a prompt.

Alibaba's text-to-video model. Cinematic quality, up to 15s.

A sophisticated, refined male voice for high‑end presentations.

Performance-optimized text-to-image with fast generation.

17B parameter open-source model with state-of-the-art quality and speed.

1T parameter omni-modal model with vision, reasoning, and agentic workflows. Free tier.

Generate 3D models from text or images.

Second generation model for graphic design.

Fast version of V2.

Balance between speed and text quality.

Highest quality Ideogram with excellent typography.

Fast text generation under 5 seconds.

Google's latest text-to-image model with high detail.

Optimized for low latency with Imagen 3 quality.

Google's flagship text-to-image model, high quality.

Fast version of Imagen 4, slightly lower quality.

Ultra high-quality image generation from Google.

A warm, approachable male voice perfect for everyday conversation.

Always points to the latest Moonshot Kimi model (currently Kimi K2.6).

Kimi K2 with native reasoning.

Native multimodal model with state-of-the-art visual coding and agent swarm capabilities.

Trillion-parameter MoE model with 32B active parameters, optimized for long-horizon reasoning and multi-step tool use.

Ultimate quality, best for final renders.

Professional grade, 1080p resolution.

Balanced quality and speed, 720p up to 10s.

Blazing fast, high quality 1080p.

Next-gen model with optional native audio.

Kling 2.6 Pro text-to-video via WaveSpeed.

Kling O3 4K image-to-video. Maintains style from start frame.

Kling O3 4K reference-to-video. Up to 7 ref images, optional ref video.

Kling O3 4K text-to-video. Intelligent scene segmentation.

Multimodal video model with advanced audio understanding.

Kling V3.0 4K image-to-video. Start frame + optional end frame.

Kling V3.0 4K text-to-video. Up to 15s, audio included.

A firm, resolute male voice that conveys confidence and purpose.

Add motion to static images.

Free embedding model with vision-language understanding.

Leonardo.ai's all-purpose text-to-image model.

Generates short music clips up to 30 seconds. Ideal for quick prototyping.

Generates full songs up to 3 minutes with detailed structure and high‑quality vocals.

An assertive, confident female voice in Indonesian.

Advanced reasoning model from Inception.

A light, airy female voice, great for children's content.

Fast and efficient MoE model with 256K context, optimized for reasoning and coding.

Full-modal model supporting text, image, audio, and video. Native multimodal with strong reasoning.

Flagship 1T parameter model with 1M context, designed for agentic workflows and complex coding.

Focus on Asian aesthetics and speed.

Optimized for understanding and generating human-like responses.

Efficient model from MiniMax, good for conversational AI.

Creates songs up to 4 minutes. 2 free trials!

Full song generation with rich instrumentation and natural vocals.

European model with high performance and focus on privacy.

Mistral's 128B dense model with 256k context, vision, and hybrid reasoning. Great for coding and agentic workflows.

Image editing with img2img support.

Advanced image editing with multi-image input support.

Edit images with Nano Banana Pro.

Ultra high-resolution image editing with Nano Banana Pro.

Free compact model from NVIDIA, efficient for edge devices.

Free high-performance model from NVIDIA, suitable for various tasks.

A serene, spiritual female voice with a calm and measured tone.

Creative-focused model with minimal restrictions. Explore diverse artistic styles.

A calm, deliberate male speaker, excellent for educational content.

Balance between performance and speed, suitable for various tasks.

Preview model with latest features from OpenAI, ideal for testing.

Most powerful model for complex reasoning, best for intricate tasks.

High-quality image generation with strong prompt adherence.

A calm, authoritative male leader voice in Indonesian.

Create personalized avatar with 1-3 reference images.

Transform a single image into a dynamic video with motion.

AI-powered video transitions between two images.

Fast generation with good quality.

Enhanced with motion control and multi-image fusion (up to 8 images).

Community model fine-tuned for photorealism and anime.

Model from Alibaba, very economical for daily use.

Balanced performance and cost from Alibaba.

Compact model from Alibaba, good for lightweight tasks.

1M context window, mandatory chain-of-thought reasoning, and improved agentic reliability. Free during preview.

Multimodal model with excellent text rendering.

Alibaba's 7B text-to-image foundation model. Excellent photorealism and typography.

Alibaba's 7B unified image editing model. Edit existing images with text instructions.

Pro version of Qwen Image 2.0 Edit with higher quality and better detail preservation.

Vision-language Qwen3 with reasoning.

Generate stunning posters in RPS style.

Upscale images up to 4x with AI enhancement.

Legacy version, more affordable.

Standard version for fast design generation.

Professional design-focused image generation.

Generate detailed editable SVG vector graphics.

A soft‑spoken, gentle female voice in Indonesian.

Quick generation with decent quality.

Preview version, fast results.

Maximum quality preview.

Professional grade image generation.

Balanced speed and quality.

A compassionate, caring male voice in Indonesian.

A wise, mature female voice, perfect for storytelling and narration.

A cute, sweet young female voice in Indonesian.

8B parameter multimodal diffusion transformer.

Stable Diffusion XL, versatile and powerful.

Ultra-fast 4-step generation, high quality.

Lightweight model from ByteDance, efficient for daily use.

Lightweight, fast, and affordable.

Professional grade with higher detail.

Faster Pro version.

Multimodal video model with support for up to 9 reference images.

Fast video editing.

Fast video editing with Turbo quality.

Fast extension of a video from its last frame.

Video editing with high quality.

Video editing with Turbo quality.

Extend a video with higher quality.

Stable text-to-image with good quality.

ByteDance's advanced image model with aspect ratio control.

Older Seedream 4 image editing model.

ByteDance's advanced image model.

ByteDance's latest text-to-image model.

Edit images using the latest Seedream 4.5 model.

Generate a sequence of edited images with Seedream 4.5.

Latest ByteDance model with multi-step reasoning.

Model with integrated web search capabilities from Perplexity.

A sweet, pleasant young female voice perfect for audiobooks.

Original Stable Diffusion model.

Generate stickers with transparent backgrounds.

Large preview model from Arcee AI, free to use.

398B MoE reasoning model with 13B active parameters, optimized for complex reasoning and agentic workflows.

Lip-sync animation from image and audio.

Google's premium text-to-video model.

Latest Veo with optional audio generation and JSON prompt support.

Faster Veo 3, slightly lower quality but great speed.

Refined Veo 3.1 with improved prompt adherence and JSON support.

Fast version of Veo 3.1 with JSON prompt support.

A commanding, regal female voice that exudes authority.

Lightweight video model, 5 seconds 480p. Fast and economical.

Optimized Wan 2.1 for image-to-video at 480p.

Ultra-fast Wan 2.1 image-to-video at 480p via WaveSpeed.

Optimized fast image-to-video with interpolated frames.

Image-to-video with precise motion control.

Advanced text-to-video with high quality output.

Alibaba's Wan 2.6 image editing model.

Alibaba's latest Wan 2.6 text-to-video model via WaveSpeed.

Next-gen text-to-image with high quality output.

Professional grade image generation with advanced controls.

8.9B parameter model built on FLUX.1-schnell. Ultra-fast with unique visual style.

Generate consistent characters from text prompts. Ideal for storytelling and branding.

Super fast 6B parameter model, sub-second generation.

Ultra-fast 6B parameter image generation.

An outgoing, lively female voice full of energy and enthusiasm.