According to a recent LinkedIn post from Together AI, Qwen4 Max is being offered at a 30% discount through November 22. The post describes a test in which the model processed a bug report and all 344 files in the Tornado repository within a single request using a 1-million-token context window.
The post says Qwen4 Max identified the relevant code and produced a patch that passed the project’s regression tests. Tornado’s broader test suite, comprising 1,673 tests, reportedly ran without failures, while the 652,519-token request cost $0.69 at the promotional price.
For Together AI, the example may support positioning around long-context model deployment and comparatively low-cost software development workflows. If replicated across broader workloads, such economics could help attract developers and enterprise users, while also increasing competitive pressure on inference providers and coding-assistance platforms.
The result remains a single company-shared demonstration conducted during a temporary discount, rather than an independent benchmark. Investors may therefore view it as evidence of product capability and pricing strategy, but would need additional information on sustained usage, margins, customer adoption, and performance across repositories before assessing its broader financial impact.

