openJiuwen releases X-Router self-evolving model routing
10/02/2026 — 10/02, 22:01·1 sources·1 reports
Story overview
On October 2, 2026, openJiuwen announced X-Router, a self-evolving model routing technology, according to a report from QbitAI published the same day. The company's stated goal is to have every request routed to a suitable model, and to feed the results of each request back into the system so that the next choice is better and cheaper. In other words, the routing policy is meant to evolve as it is used rather than stay fixed.
openJiuwen says X-Router is Ascend-friendly, and that in its own testing it reduced Token consumption by more than 50%. Those figures come from the company itself. The report does not name the models involved, the baseline used for comparison, the tasks that were tested, or how the technology is packaged or deployed.
The report describes the release as a first for openJiuwen, matching the "first release" wording in its headline, and it is so far the only coverage of the announcement. Beyond the company's own description of how X-Router works and its claim of cutting Token use by over 50%, no third-party verification and no further developments have been reported, so the matter currently rests at the point of announcement.
AI-generated from 1 reports · updated 2 hours ago
Latest turnopenJiuwen launched X-Router, a self-evolving model routing technique that picks the right model per request and feeds each result back into later routing decisions. The company says it is Ascend-friendly and cut token consumption by more than 50% in its own tests.

- Heat index
- 77
- Sources
- 1
- Reports
- 1
- First seen
- 9 hours ago
Reports on this story headlines open the original
openJiuwen launched X-Router, a self-evolving model routing technique that picks the right model per request and feeds each result back into later routing decisions. The company says it is Ascend-friendly and cut token consumption by more than 50% in its own tests.
量子位AI score 72
Other stories people are talking about
- 292Google releases Gemini 4 Argon as next-gen frontier AI model9 sources
- 181NewNVIDIA launches 64GB DGX Spark desktop AI computer at $4,9992 sources
- 101AWS launches multiple AI decision and agent products2 sources
- 98NewAWS fine-tunes search agents with multi-turn RL on SageMaker AI1 sources
- 98NewAWS adds web search to Claude Desktop via Bedrock AgentCore1 sources
- 98NewAWS pairs Amazon Quick with an MCP rules engine for lease compliance1 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.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion
