Primate Labs is releasing Geekbench 7, the latest version of its popular tool for measuring how fast computers and phones are. The update adds new audio and video tests, a redesigned multi-core test, and larger datasets for both CPU and GPU evaluation. The release was announced on July 23, 2026 The Verge.

A benchmark is a program that runs a series of tasks on a device and produces a score, similar to how a driving test measures a car's handling, braking, and acceleration under controlled conditions. Geekbench runs on Windows, Mac, Linux, Android, and iOS, which makes it easy to compare very different devices on the same scale.

The most notable addition is a set of audio and video tests that Primate Labs modeled around things people actually do: video calls, streaming, and watching media. These tests include AV1 and Opus, which are modern, royalty-free formats for video and audio. AV1 handles video; Opus handles audio. Both are open standards, meaning device makers and software developers can use them without paying licensing fees.

AV1 has been adopted widely by streaming platforms and web browsers over the past several years, and Opus is now the default audio format in many video calling apps. Their inclusion in Geekbench 7 means the tool now tests the same formats that people use every day, rather than older or more artificial substitutes.

The multi-core test has also been redesigned. Most modern chips have several processing cores, which are like workers on an assembly line. Previous versions of Geekbench split every task across all available cores to get the highest score. Geekbench 7 takes a different approach: it only splits tasks across multiple cores when the real-world version of that task is actually designed to use multiple cores. The practical effect is that scores should better reflect how a chip performs on tasks that genuinely benefit from multiple cores, rather than simply rewarding chips that have more of them. Primate Labs has published an official CPU Workloads documentation PDF on geekbench.com that details the benchmark's test categories, including data compression tests among other workload types Primate Labs.

Both CPU and GPU tests in Geekbench 7 use larger, more demanding datasets than the previous version. Larger datasets place greater pressure on a chip's cache (fast, built-in memory) and memory bandwidth (how quickly data moves in and out of the chip), which is where many modern devices, particularly those with integrated graphics, differentiate themselves.

Geekbench 7 scores are not compatible with Geekbench 6 scores. Users comparing results across the two versions will need to re-run the benchmark on Geekbench 7 to establish a new baseline. The Geekbench Browser, Primate Labs' public results database, will presumably accumulate Geekbench 7 results alongside the existing archive of prior-version scores, though cross-version comparisons will not be valid.

A Geekbench 7 Pro license is priced at $99, with an introductory discount bringing it to $79 through August 6th. The Pro license covers Windows, macOS, and Linux desktop platforms and enables offline benchmarking, meaning results are not automatically uploaded to the Geekbench Browser. For businesses, hardware reviewers, and security-conscious organizations that need to test systems without exposing results publicly, the offline mode addresses a practical constraint that has existed since the browser-upload model became the default behavior.

The broader context here is how Geekbench 7's design choices reflect a shift in what matters for system performance. The inclusion of AV1 and Opus workloads signals that media performance, much of it now handled by dedicated hardware on modern chips, has become a first-order concern for users. A chip that posts strong general-purpose scores but struggles with AV1 video playback will deliver a poor experience in exactly the scenarios Primate Labs is now modeling: joining a video call, watching a stream, or consuming media.

The redesigned multi-core test is arguably the more consequential change for the enthusiast and professional audience. For years, benchmark suites have grappled with the tension between artificial parallelism, which rewards core count, and realistic parallelism, which reflects how software actually behaves. By limiting multi-core testing to workloads that genuinely use multiple cores in practice, Primate Labs is making a statement about what multi-core performance should mean. Whether the industry accepts that framing or continues to lean on more aggressively parallelized metrics will be worth watching as results populate the Geekbench Browser in the coming weeks.

What this enables, ultimately, is a benchmark suite better aligned with the tasks that dominate everyday computing in 2026. The tradeoff is a clean break from prior-version comparability and, for Pro users, a license fee. For hardware reviewers and system architects who rely on Geekbench as one data point among many, re-establishing baselines on Geekbench 7 is a straightforward if necessary step.