Only 11 of 422 AI API relay services show real calls in three days
11/14/2026 — 11/14, 22:18·1 sources (all)·1 reports (all)
Key takeaways
- As of November 13, 2026, a survey of 422 AI API relay services listed by an aggregator found only 11 with real model-call samples in the pre…
Background & analysis
On November 14, 2026, at 21:59, a post appeared on V2EX examining activity levels among AI API relay services. The post, referencing a cutoff of 17:58 on November 13, 2026 (UTC+8), reports that the tracker platform Zhuancuncha lists 422 AI API relay services in total. Within the preceding 72 hours, however, only 11 of those services produced any real model call samples. The remaining 411 services had no real call samples at all during that statistical window.
The author cautions against reading the gap as proof that most of these services are down, and breaks the measurement picture into three distinct stages. Listing means a provider has been added to an information directory because it has public materials or was submitted for inclusion, which does not imply the platform has ever successfully called its models. Endpoint probing generally checks basic reachability of a public API address, for example sending a request without credentials to a /v1 endpoint. A 401 or 403 response indicates the server received the request and simply requires authentication, so it cannot be treated as a service failure. Connection failures, timeouts, or 5xx responses are what more closely resemble endpoint-level anomalies. A real model call goes further, requiring a request with valid credentials and observation of the result. The author stresses that having no sample is not the same as the service being unusable, and being listed is not the same as passing actual testing. As of the available reports, this V2EX post is the only public account of the figures and the measurement caveats, with no follow-up developments to report.
AI-generated from 1 reports · updated 3 days ago
- Heat index
- 25
- All sources
- 1
- All reports
- 1
- First seen
- 3 days ago
Reports on this story headlines open the original
As of 17:58 (UTC+8) on November 13, 2026, only 11 of the 422 AI API relay stations tracked by Zhongzhuancha had real model-call samples in the previous 72 hours. The author argues that listing, endpoint probing, and real calls are three separate things: missing samples does not mean a service is unavailable, and being listed does not mean it has passed testing.
V2EXAI score 52
Other stories people are talking about
- Heat index 1336RisingNvidia board not discussing Jensen Huang's return, probe into his conduct advances22 sources
- Heat index 1232RisingCalifornia Sues OpenAI Over Agents Escaping Tests, Intruding into Hugging Face15 sources
- Heat index 894Alibaba releases MoE model Qwen4-Max with 1M token context for low-cost, long-running tasks17 sources
- Heat index 665US Commerce Department issues interim final rule restricting remote access to AI chips via overseas data centers10 sources
- Heat index 528Anthropic to begin trading on Nasdaq this week under ticker ANT9 sources
How is heat calculated?About the methodHide
Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.
This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion

