Bytespider
Bytespider is a training crawler operated by ByteDance. Its job is to read public web pages and feed that content into a machine-learning pipeline that trains future versions of the model.
Unlike a search-engine crawler, a training crawler does not send users back to your site. The content is consumed once, baked into the model, and shows up later in the model's responses. There is usually no citation and no referral traffic.
If ByteDance ships a new model version, you will likely see Bytespider traffic spike for a few weeks while it gathers fresh data, then quiet down again.
See Bytespider on your own site
Match the User-Agent header on incoming requests against the pattern below.
regex
Known UA examples
example 1
For higher confidence, also verify the source IP against the operator's published ranges. UA strings can be spoofed; IP ownership is harder to fake.
Renders JavaScript
No
IP verification
Published IP ranges
Crawl frequency
Periodic, broad
Honors robots.txt
Yes
Honors Crawl-delay
No
ByteDance runs 8 bots in total. Each one is a separate user-agent so you can allow or block them independently.
Training Crawler
3- BytespiderYou are here
- imageSpider
- DoubaoBot
Link Unfurler
1AI Coding Tool
1Search Engine
1Live-Fetch AI
1Shopping Bot
1Share of AI bot traffic
8.8%
↓ 29% vs 12 mo ago
% of top sites blocking it
89%
↑ 9% vs 12 mo ago
Top page categories crawled
Training Crawlers as a category, share of all bot traffic
68%
↓ 21% vs 12 mo ago
Sum of every training crawler we have public stats on. Compare to Bytespider's individual trend above to see if it's pulling ahead of its category or falling behind.
Crawl-to-click ratio
76,000 : 1
For every 1,000 fetches, this agent sends ~0.013 human visits back. Cloudflare crawl-to-click data
Snapshot updated
2026-04
Connect Agent Analytics for daily, per-site numbers.
Source: Cloudflare Radar AI Insights
Should I let Bytespider through?
There's a real trade-off here. Training crawlers consume content without sending users back. Decide whether the trade for being in the model is worth your bandwidth. If ByteDance actually drives traffic or citations back to you, letting it through usually pays for itself. If it just consumes bandwidth, block it.
Does blocking Bytespider affect my Google rankings?
No. Bytespider collects training data, not search-index pages. Your classical search rankings stay intact. The actual trade is whether you want your content folded into the next model release.
How do I confirm a request is really from Bytespider?
Look at the User-Agent header in your access logs and match it against the strings listed above. Worth knowing that the User-Agent is easy to fake, so this check tells you "the traffic claims to be Bytespider", not "the traffic is genuinely Bytespider". If you need stronger guarantees, look for a reverse-DNS check or wait for ByteDance to publish IP ranges.
What happens to my content if I let Bytespider fetch it?
It gets pulled into ByteDance's training pipeline and stored. Whether and how it influences a future model release is rarely disclosed. The only real lever you have on the outcome is what you allow at fetch time.
How is Bytespider different from ByteDance's other bots?
ByteDance splits work across multiple user-agents so site owners can decide on each one independently. Training crawlers, live-fetch agents, search indexers, and agentic browsers each get their own name. Worth scanning the rest of the ByteDance family above to see which ones actually matter for your site.
What's the cleanest way to control Bytespider?
Two layers. Robots.txt for the polite crawlers that read it, and rules at your CDN or edge for the ones that don't. Rankly's Agent Experience handles both from a single config, so you can allow, block, rate-limit, or serve a stripped-down version per bot. Agent Analytics handles the observation half so you know which bots are actually worth a rule.
Verify everything above against the operator's own documentation.