To find where an image really came from, run it through several reverse image search engines, sort or filter the matches by date to find the earliest copy, then confirm that copy with archives, the uploader's history and the image's own content. No single tool knows when an image first appeared online, so the goal is to build a timeline from several pieces of evidence.
Key takeaways
- Search with at least two engines. Google Lens, TinEye and Bing index different parts of the web.
- TinEye's Sort by: Oldest shows the copies its crawler found first. That is a crawl date, not a publication date, but it is an excellent lead.
- The Internet Archive's Wayback Machine can prove a page and image existed by a certain date.
- The largest, least-edited, uncropped version is usually closest to the original. Watermarks, credits and captions on that copy point to the creator.
- "Earliest copy we can find" is not always "the original". Say how confident you are and why.
Why "original source" is harder than it sounds
Images are copied, cropped, filtered, screenshotted and re-uploaded constantly. Each copy can lose its caption, credit and metadata. Search engines also don't record when an image was first published; they record when their crawler found it. TinEye's own help pages make this point: it can tell you when TinEye first found an image, not when it first appeared online.
So a good answer usually takes this form: "The earliest copy we can find was posted by this account on this date, it is the highest-quality version, and it carries the photographer's watermark." That is strong evidence, but it is still a conclusion drawn from evidence. Keep that framing and you'll avoid most mistakes.
Prepare the image
- Get the best copy available. Open the image in its own tab and save that, rather than a thumbnail. On social media, tap through to full size first.
- Note everything around it. Record the URL, account name, caption, post date and any visible credit before you go further. Posts get deleted.
- Make variants if needed. Crop away added text, borders, stickers or interface elements. Crop to the most distinctive object or person. If you suspect a mirrored copy, flip it horizontally and search that too.
- For video, use a frame. Pause on a sharp, distinctive frame and screenshot it. Our reverse video search guide covers keyframe tools.
Search several engines
Each engine has strengths, so run the same image through more than one. Our reverse image search guide has full instructions for each platform, and this comparison covers more tools.
| Tool | Best for source-hunting | What to click |
|---|---|---|
| Google Lens | Largest index; news and fact-check context | Exact matches to drop look-alikes; About this image in Lens and Circle to Search for how sites describe it |
| TinEye | Exact and edited copies, with crawl dates | Sort by menu, then Oldest; Compare to spot edits |
| Bing Visual Search | Different crawl coverage; good for some social and shopping pages | The "pages with this image" style results list |
| Yandex Images | Often strong on faces and on Eastern European and Russian-language sites | Sites list under the image matches |
On Google, uploading an image now opens Lens by default. Lens leans toward "visually similar" results, so switch to the exact-match view when you want copies of this specific photo. In Chrome you can also right-click any image and choose Search with Google Lens. If TinEye isn't finding enough, our TinEye alternatives page lists other options.
Narrow down by date
Once you have a pile of matches, you want the oldest ones.
TinEye: sort by Oldest
On a TinEye results page, open the Sort by menu and choose Oldest. The copies TinEye found earliest move to the top. The other sorts are useful too: Biggest image often surfaces the highest-resolution copy (likely closer to the original), and Most changed shows heavily edited versions, which can reveal manipulation. TinEye hides results that are no longer online by default; you can choose to include them, which helps when the first poster deleted their copy.
Google: date filters and operators
For keyword searches about the image, open Tools under the search box, then the Any time menu, and pick Custom range. You can also type before:2023-06-01 or after:2023-01-01 into the query. Google's Search Liaison has described these operators as still in beta and noted that Google can't always know a page's true date, so treat results as leads.
Check dates on the page itself
Every promising match needs its own date check. Look at the post timestamp, the article byline date, and comments with dates. Be wary of pages that show today's date or "updated" dates, and of aggregator sites that republish old content under new dates.
Confirm with the Wayback Machine
The Internet Archive's Wayback Machine at web.archive.org stores snapshots of web pages over time, giving you independent, timestamped evidence.
- Paste the URL of the page where you found an early copy and click Browse History.
- Use the calendar view. Blue dots are successful captures; pick the earliest one.
- Open the snapshot and check whether the image is there. If it shows as broken or gray, the image itself may not have been archived.
- You can also search the image file's own URL to see when that file was first captured.
- To preserve a page you found, use Save Page Now, which captures that one page on demand.
The 14-digit number in an archive URL is the capture time in the format yyyymmddhhmmss. Gaps are normal: sites can block the crawler, and pages behind logins (much of social media) are often missing. A capture proves the image existed by that time; it doesn't prove that page was first.
Check the uploader
When you have a candidate original, look at the account or site that posted it:
- Do they create this kind of content? A photographer's portfolio with dozens of shots from the same shoot, including near-identical frames, is strong evidence. An account that only reposts viral content is not.
- Are there companion images? Other photos taken seconds apart, wider shots or behind-the-scenes images are hard for a copier to have.
- Does the account's location and history fit? Someone claiming to photograph an event should plausibly have been there.
- Can you contact them? Photographers and journalists will often confirm their work, and may share the original file. If you plan to reuse the image, this is also how you get permission.
Use metadata and content clues
If you can get the original file, its embedded metadata may show the camera model, capture time and sometimes GPS coordinates, plus creator and copyright fields. Most social platforms strip this data on upload, and anyone can edit it, so treat it as supporting evidence. Our photo metadata guide explains how to read it.
The image content also carries clues: shop signs and their language, vehicle models and plates, weather, shadows that indicate time of day, and landmarks you can find on a map. Comparing these with what is claimed often exposes "wrong place, wrong date" recycling, which is one of the most common forms of photo misinformation.
Verifying viral and news photos
Breaking news brings a flood of recycled images: photos from earlier disasters, other countries, films or video games, and increasingly AI-generated images. Before you share:
- Search for existing fact-checks. Lens and About this image often surface fact-checking articles. You can also search the claim plus "fact check".
- Look for corroboration. Real events produce many images from different people and angles. Check reputable news outlets and wire services.
- Check that the scene fits. Weather, season, signage language and architecture should match the claimed place and date.
- Consider AI. If the image has no earlier copies and looks too dramatic or perfect, run the checks in our guide to spotting AI-generated images.
Fact-checkers widely use the free InVID-WeVerify browser plugin, which bundles reverse search across several engines, a metadata reader and a magnifier into one toolbar.
A worked example (hypothetical)
The claim: A post shared widely this week shows a flooded city street with the caption "This is downtown right now."
- Prepare. You save the full-size image and note the posting account, which was created two months ago. You crop out a caption banner that was added on top.
- Search. Google Lens exact matches show dozens of reposts from the last two days. TinEye finds 40 matches.
- Sort by date. With TinEye set to Oldest, the top result is a regional news site's gallery found by TinEye several years earlier. Biggest image points to a larger copy on a photographer's portfolio site.
- Confirm. The Wayback Machine has a capture of the news gallery page from the same year, with the image visible and a caption naming a different city and a storm from that year.
- Check the uploader. The portfolio contains 15 other frames from the same street that day, and the news gallery credits the same photographer.
- Clues. Shop signs in the photo are in a language not commonly used in the city named in the viral caption.
- Conclusion. High confidence: the photo is real but recycled, taken years earlier in another city by a credited photographer. The viral caption is false.
Notice the conclusion rests on several independent pieces of evidence. TinEye's date alone would only have been a lead.
Frequently asked questions
Can I find out exactly when a photo was first posted online?
Usually not with certainty. Search engines record when they found a copy, not when it was published, and many first posts are deleted or on platforms crawlers can't reach. You can establish the earliest known copy and back it up with timestamps and archive captures.
What if the image has no matches anywhere?
Try cropped and flipped versions, a different engine, and keyword searches describing the scene. If nothing turns up, the image may be new, private, heavily edited or AI-generated. That is "unknown", not "original".
Is the highest-resolution copy always the original?
Often, but not always. Upscaling tools can enlarge a copy, and stock sites may host larger versions than the first post. Combine size with dates, uncropped framing and the uploader's history.
I found my own photo being used without permission. What now?
Save screenshots and URLs as evidence, then contact the site or use its copyright reporting process. Our guide to DMCA takedowns for stolen photos walks through it.
Sources & further reading
- TinEye Help: Can I sort my results?
- TinEye Help: TinEye tutorial
- Google Search Help: Search with an image
- Google: New ways to access About this image
- Internet Archive Help: Using the Wayback Machine
Spotted an error or something out of date? Tell us — we review corrections promptly. Read how we research and update guides in our editorial policy.