Meta делает конкурента OpenRouter: маршрутизатор AI-задач ради экономии на вычислениях

Meta's internal lab called AAILabs is developing an AI-task routing service — an analog of OpenRouter — that distributes requests among more cost-efficient and larger models to cut the company's compute costs. This was reported in July 2026 by 36Kr, citing internal documents reviewed by Sina Finance (新浪财经).

What AAILabs Is Developing

AAILabs is building an internal dispatcher service that routes different AI tasks to the large models with the best price-to-quality ratio. According to 36Kr, the key goal is to reduce overall compute spending, including in code development: routine requests go to cheaper models instead of running everything through the most expensive ones.

*AAILabs is an incubator within Meta's Applied AI engineering team

*The department was officially created in March 2026

*The service is similar to OpenRouter — it routes tasks among different LLMs

*The goal is to cut compute costs, including for code development

*Source — 36Kr, citing Sina Finance (新浪财经)

How Meta's Incubator Works

Inside AAILabs, Meta employees submit proposals for AI products and services for internal use within the company. The incubator belongs to the Applied AI engineering team and, according to 36Kr, was formed in March 2026 specifically as a platform for such initiatives.

Once a proposal is approved, a small dedicated team is formed to bring the product to completion. According to internal documents, this same stage is when it's assessed whether the service should be released beyond the company — some mature projects may get a full public release.

"After a project is approved, the company assembles a small dedicated team to finish developing the product and assesses whether to release it externally," — from

Meta's internal documents reviewed by Sina Finance.

Why Meta Wants Its Own OpenRouter

Meta is building its own model router because compute bills for working with large language models grow along with the number of internal AI tools. OpenRouter, which the project is modeled on, is a gateway that distributes requests among dozens of LLMs from different providers; a similar internal system allows automatically choosing, for each task, the model with the best price-to-result ratio.

For Meta, this is primarily a matter of cost savings: the more code and text employees generate via AI, the more routing between cheap and expensive models affects the final infrastructure budget.

What This Means

Large tech companies are starting to treat AI compute as a cost line that needs to be optimized, not just scaled up. Meta's internal OpenRouter rival is a signal that routing tasks among models by price is becoming standard practice for cutting costs at scale.

Frequently Asked Questions

What is AAILabs developing?

AAILabs, Meta's internal incubator, is building a routing service — an analog of OpenRouter that distributes different AI tasks to more cost-efficient large models to cut the company's compute costs.

When did Meta create AAILabs?

According to 36Kr, the department was officially formed in March 2026 and belongs to the Applied AI engineering team within Meta.

Will the service be released outside Meta?

According to internal documents, mature AAILabs projects may get a public release: after development, a dedicated team assesses whether to release the product externally.

*Meta has been designated an extremist organization and is banned in Russia.