File size: 1,968 Bytes
6208234
 
018c285
58d7456
 
6208234
 
 
 
25cf0d9
 
018c285
 
 
 
 
 
 
25cf0d9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
---
title: README
emoji: 🟪
colorFrom: pink
colorTo: purple
sdk: static
pinned: false
---

# Everypixel

Everypixel builds generative media and computer vision tools. We also run blind benchmarks of AI models.

## Everypixel app

A web workspace for image, video, voice, and lipsync generation. It gives you GPT Image 2, FLUX.2, Seedream 4.0, Kling 3.0, Seedance 2.0, Runway Gen 4.5, WAN 2.7, and other leading models in one account. Every result is saved to a shared asset library. The free plan includes 200 credits, no card required.

[Open Everypixel](https://app.everypixel.com/)

## Everypixel Labs API

Pay-as-you-go endpoints with no subscription. You get 500 free requests, no card required.

| Endpoint | What it does | Price |
|---|---|---|
| `/image_keywording` | Image tagging in 7 languages, plus color recognition | $0.6 / 1,000 requests |
| `/image_captioning` | Detailed alt text from an image | $1.2 / 1,000 requests |
| `/photo_quality` | Scores and ranks images by quality | $0.6 / 1,000 requests |
| `/age_recognition` | Estimates a person's age from a face | $0.6 / 1,000 requests |
| `/text_to_speech` | Converts text to natural speech | $0.02 / 60 sec |
| `/lipsync` | Syncs a voiceover with video | $1 / 60 sec |
| `/face_swap` | Swaps faces in images and video | $10 / 1,000 requests |

[API overview](https://labs.everypixel.com/api) · [Docs](https://labs.everypixel.com/docs)

## Everypixel Research

Blind pairwise evaluations of image generation and upscaling models, judged by working professionals. Recent tests:

- **GPT Image 2 vs real photography:** GPT Image 2, Qwen Image 2, and FLUX.2 Klein 9B against real photos. 48 scenes, 13 professional judges.
- **SeedVR2 vs Magnific:** 264 image pairs, 13 evaluators, 792 judgments. SeedVR2 led on detail recovery (86%) and naturalness (82%).
- **SeedVR2 vs Topaz Wonder 3:** 264 pairs. SeedVR2 was stronger on edges, Topaz on faces.

[All benchmarks](https://research.everypixel.com/)