flux-1-kontext-dev
Edits existing images using text instructions, allowing for changes in style, characters, or objects, and reliably handles multiple edits while maintaining image coherence.
image
bytedance-uso
A unified image editor that allows users to generate images by combining any subject with any style efficiently, preserving identity and consistency.
imageimage
32
image-resize
resize images by width, height, scale factor, or megapixel target
hidream-i1-full
generates high-quality images with state-of-the-art results.
flux-1-kontext-dev
edits existing images using text instructions, allowing for changes in style, characters, or objects, and reliably handles multiple edits while maintaining image coherence.
sdxl
generates and modifies images from text prompts, producing high-resolution, photorealistic results with superior detail and accuracy compared to previous versions.
flux-1-fill
fills designated areas in existing images based on a descriptive text input.
hidream-e1-full
edit images by simply telling it what changes you want, such as altering colors, backgrounds, or accessories with precision.
real-esrgan
enhances low-quality, degraded images and videos by upscaling resolution, reducing noise, and restoring fine details.
qwen-image-edit-lightning-plus
edit images using text prompts or by incorporating elements from multiple source images, making it easy to add subjects or put several items into one scene.
hidream-i1
an open-source tool for generating high-quality images in seconds.
thera
upscales images to any size without blurriness or jagged edges, maintaining high detail through its unique neural heat field technology.
flux-1-dev-upscaler
increases the resolution of images and videos, enhancing clarity and detail by up to 4x.
falconsai-nsfw-detection
detects nsfw content in images and videos using falconsai/nsfw_image_detection model. for videos, samples frames at configurable intervals.
z-image
a fast and high quality image generation model
sdxl-controlnet
generates images with high quality and precise control over the composition and structure.
qwen-image-edit-plus
performs high-quality editing across multiple images.
cogview4-6b
generates high-quality images from text, capable of producing detailed visuals up to 2048x2048 resolution.
bytedance-uso
a unified image editor that allows users to generate images by combining any subject with any style efficiently, preserving identity and consistency.
flux-1-srpo-dev
creates high-quality images from text descriptions, specializing in photorealistic results.
flux-1-krea-dev
a tool for generating images from text descriptions, available for non-commercial use.
wan2-2-i2i-a14b
creates videos from images and enhances video quality using a built-in upscaler.
qwen-image
generates and edits images from text descriptions, excelling at rendering complex text within the image.
qwen-image-edit
advanced image editing that excels at rendering and manipulating text within images, allowing for precise changes to appearance and meaning.
omni-zero
creates stylized portraits instantly without needing specific training data.
sdxl-inpainting
fills in missing or masked areas of an image using text instructions to guide the replacement content.
omni-try
try on clothing and accessories virtually before buying.
birefnet
removes backgrounds from images quickly and accurately.
z-image-turbo
a fast and high quality image generation model
flux-1-dev-controlnet-union
creates and modifies images using multiple structural control techniques like edge detection, depth, and pose in a single tool.
hidream-i1-fast
a rapid image generator that produces high-quality images in various styles very quickly.
chroma-1-hd
creates high-definition images from text descriptions, primarily designed as a robust base for developers to customize and build upon.
sd-inpainting
fill in or replace masked parts of an image using a text description.
stitch-images
combine multiple photos horizontally or vertically into a single image or collage.
video
20
html-to-video
render html/css/js animations to video — supports gsap timelines, css animations, web animations api
hyperframes-render
render heygen hyperframes compositions to video — supports clips, gsap timelines, track layering
remotion-render
render videos from react/remotion component code — pass tsx, get mp4
videoseal
embed or detect imperceptible watermarking for images and videos.
wan2gp-i2v
turns static images into dynamic, realistic videos, often including tools for advanced control over motion and video style.
wan2-1-i2v
generates high-quality, dynamic videos from a single static image.
rife-video-interpolation
increases video smoothness by using ai to generate and insert extra frames, effectively raising the video's frame rate.
extract-last-frame
save a specific frame from the end of a video as a static image file.
wan2-2-t2v-5b
generates video content from text prompts.
media-merger
merge, concatenate, and stitch multiple videos and images into a single video with customizable transitions. supports common formats (mp4, mov, webm, png, jpg), custom frame rates, and transition effects between clips. built on ffmpeg
latentsync-1-6
generates precise and realistic lip synchronization for video and audio content, ensuring high-resolution and smooth temporal consistency.
wan2-2-i2v-a14b
generates high-resolution videos from single images, offering one of the fastest ways to create 720p, 24fps video clips from a still picture.
ltx-video
create high-quality, realistic, and customizable videos quickly, with capabilities for producing detailed, high-resolution content.
wan2-1-t2v-effects
generates high-quality videos from text descriptions with dynamic motion and supports both chinese and english language prompts at 720p resolution.
wan2-2-i2v-5b
generates high-definition video from images or text prompts quickly, running efficiently even on standard graphics cards.
ovi-ti2v
creates videos with synchronized audio from text descriptions or images, supporting both text-to-video and image-to-video generation with built-in speech.
wan2-2-t2v-a14b
generates video content using text prompts.
wan2-1-i2v-effects
generate videos from text descriptions and apply image-to-video effects for enhanced visual results.
wan2-2-i2v-5b-turbo
quickly converts still images into high-definition videos (up to 720p) at 24 frames per second, using specialized optimization for very fast results.
latentsync-1-5
generates realistic, high-resolution video where the lips perfectly match the accompanying audio input, useful for tasks like dubbing or creating virtual avatars.
other
19
agent-browser
browser automation for ai agents. navigate, interact with @e refs, take screenshots, record video with cursor indicator, execute javascript. supports proxy configuration.
flux-2-klein
the flux.2 [klein] is the fastest image models of the flux.2 family. it unifies generation and editing in a single compact architecture, delivering state-of-the-art quality with end-to-end inference in as low as under a second. built for applications that require real-time image generation without sacrificing quality.
ltx-video-2
ltx 2.0 audio-video foundation model. generates videos with synced audio. supports t2v, i2v, long video generation with multi-prompt sliding windows, and lora adapters.
caption-videos
add captions to videos using an existing caption file, such as those generated by a speech-to-text service.
array-switch
allows you to choose between two different inputs based on a condition applied to an array of data.
array-element-switch
selects one of two possible inputs based on a comparison check within an array.
mask-image
combines two images—a main image and a semi-transparent mask—to selectively hide or reveal parts of the main image, creating a partially transparent result.
video-audio-merger
merge video and audio files easily, with the flexibility to keep the original audio from the video.
text-to-file
creates a new document using the text and file name you specify.
boolean-switch
selects one of two possible inputs based on whether a condition is true or false.
numerical-switch
selects one of two inputs based on a condition involving numerical comparison.
get-item-in-list
returns a specific element from a list based on its position.
html-to-image
turns web content into customizable png or jpeg images.
extract-media-duration
extracts the length of video and audio files.
text-templating
dynamically generate content by combining a fixed template with specific data inputs.
video-audio-extractor
extracts audio from video files and removes the original audio to create silent videos.
string-switch
compares strings to decide which of two inputs to use.
text-split
splits a piece of text into smaller sections based on a specified separating character or phrase.
bounce-repeat-videos
repeats a video segment by playing it forward and then immediately backward to create a bouncing, looping effect.
audio
10
omnivoice
zero-shot text-to-speech with voice cloning for 600+ languages.
vibevoice
generates high-quality podcasts using natural speech and voices for multiple speakers.
tencent-song-generation
generates high-quality songs, including both music and vocals, and can also produce pure music or vocal-only tracks.
dia-tts
dia tts - generate realistic dialogue with emotion control, natural nonverbals, and voice cloning
diffrythm
generates complete songs quickly and simply using advanced latent diffusion technology.
audio-x
generates audio from any input using a unified framework.
mmaudio
transforms silent videos into immersive experiences by generating high-quality, perfectly synchronized sound effects and environmental audio based on the visual content.
kokoro-tts
converts text into spoken audio.
higgs-audio
generates speech from text with advanced, expressive audio quality.
chatterbox
converts written text into spoken audio.
chat
10
qwen3-30b-a3b
a powerful language application that excels at multilingual communication and complex task execution, designed for fast performance.
gemma-3-12b-it
developed by google, this open-source tool processes both text and images to answer questions, summarize content, perform reasoning tasks, and understand images.
xlam-2-32b-fc-r-i1
a system capable of advanced multi-step reasoning and a strong understanding of language and context to create actionable plans.
phi-4-14b
a powerful tool developed by microsoft and trained on high-quality data to excel at complex tasks like advanced math, coding, and general problem-solving, offering detailed reasoning alongside solutions.
devstral-small-2505
an agent for software engineering tasks, created by mistral ai and all hands ai.
mistral-small-3-2-24b-it-2506
follows precise instructions, excels at function/tool calling, and can process both text and images for tasks like document understanding and content generation.
gemma-3-27b-it
handles complex tasks like question answering, summarizing, and reasoning across both text and image inputs, with support for multiple languages.
gemma-3n-e4b-it
a fast and versatile tool that can analyze and respond to information from text, images, and audio, designed to run efficiently on small or limited devices.
phi-4
a powerful language model that excels at understanding and generating text across more than 20 languages. it is particularly effective for tasks like summarizing content, answering questions, translating, and interpreting both audio and images, including charts and tables.
magistral-small-2506
a high-performance system for complex reasoning, coding, and math problems, featuring strong multilingual support.
text
5
harrier-oss-v1
multilingual text embedding using microsoft's harrier oss v1 models. supports retrieval, clustering, semantic similarity, classification, and reranking with state-of-the-art mteb v2 scores.
deepseek-ocr-2
next-gen document ocr with improved math, tables, and reading order. converts images and pdfs to structured markdown.
python-executor
runs and executes python programming code in a safe environment.
search-assistant
helps users create and refine search queries, retrieve relevant results from various sources, and generate overviews or summaries of the information found.
deepseek-ocr
extracts and converts text from documents, especially good at reading images, photos, handwriting, charts, and graphs.
3d
2
explore more on inference.sh
infsh is one of many providers on the grid. discover hundreds of apps across image, video, audio, and more.
we use cookies
we use cookies to ensure you get the best experience on our website. for more information on how we use cookies, please see our cookie policy.
by clicking "accept", you agree to our use of cookies.
learn more.