Independent/Reader-funded/Infrastructure, not tokens
DeAINEWS

AI you control — open models, private inference, and the networks that run them.

PulseOpen-Weights Releases

Claim: open-weight models now take 72% of OpenRouter token share

An X post claims open-weight models carry 72% of OpenRouter token volume. Unverified, but OpenRouter's public rankings page can confirm or falsify it today.

DeAI is powered by Morpheus (mor.org). We cover competing providers on the same terms — see our methodology.

A single network switch with a dense bundle of lit ethernet cables plugged into it, one port glowing amber, mounted in a dark rack — the routing layer where token volume concentrates. Illustration: DeAI
A single network switch with a dense bundle of lit ethernet cables plugged into it, one port glowing amber, mounted in a dark rack — the routing layer where token volume concentrates. Illustration: DeAI

The claim making rounds today: open-weight models now carry 72% of tokens on OpenRouter. It comes from a single X post with no link attached, and the reason it matters is that OpenRouter's own public rankings page can settle it with one dated capture.

Key facts

What's driving the conversation

The post from @LeaderX_btc pairs the 72% figure with a familiar argument: open-weight models are eating routing volume, distillation lets closed labs commoditize their own frontier models, and black-box providers carry concentration risk. Medium velocity. The number is exactly the kind of datapoint the week's larger inference-economics thread from Deedy Das leans on, which is why it is spreading: if open models dominate routed traffic and provider margins are fat, the serving land-grab follows.

Replies are the usual split — people asking for the source, people posting screenshots of OpenRouter's leaderboard that neither confirm nor date the 72% figure.

The substance

What would verify: OpenRouter's public rankings and token-share page, captured with a date, showing open-weight models' aggregate share. The page publishes the underlying data, which makes this one of the rare X claims with a one-click verification path. What stays claim-only until then: the 72% number itself. Do not repeat it as fact — our DeepInfra and router coverage has shown before that adoption narratives outrun the leaderboard.

Why builders are watching

If roughly accurate, the figure says open models are the default on multi-model routers, not the exception — which changes how builders think about portability, pricing pressure, and the alternatives to any single router. It also feeds the privacy conversation: router volume means prompts flow through an intermediary, which is central to what OpenRouter logs and how routers compare on our coverage. Watch for a dated rankings capture this week; either it confirms the share or the claim dies quietly.

Questions

Do open-weight models really handle 72% of OpenRouter tokens?
That is a claim from a single X post with no link attached, and DeAI has not verified it. OpenRouter publishes public model rankings and token-share data, so the figure can be confirmed or falsified against a dated capture of that page. Treat 72% as a claim until then.
Why does OpenRouter token share matter?
OpenRouter is one of the largest multi-model inference routers, so its token distribution is a proxy for which models builders actually route production traffic to. A high open-weight share means open models are the default on routers rather than the exception.

Sources

  1. X post by @LeaderX_btc claiming 72% open-weight token share — X
  2. OpenRouter public model rankings and token share — OpenRouter

About DeAI

DeAI is an independent publication covering open-weight AI models, private inference, and decentralized infrastructure — the tools for running AI you actually control. We test providers on price, privacy, and refusal behavior and publish the numbers, not the vibes. DeAI is powered by Morpheus (mor.org), a decentralized inference marketplace, and covers it on the same terms as every other provider.

Powered by Morpheus and StrandCMS

Morpheus is a decentralized inference marketplace, covered on the same terms as every other provider — we rank it wherever the data lands. StrandCMS is the open-source, agent-first framework this site is built on.

Learn more about the Morpheus Inference API →

Sponsor disclosure — not editorial

Powered by Morpheus and StrandCMS. Morpheus is a decentralized inference marketplace, covered on the same terms as every other provider. StrandCMS is the open-source, agent-first framework this site is built on.

Learn more →