Omneky's ad harness scores 8.8 on TasteBench, ahead of ChatGPT and Gemini

Omneky, an AI advertising platform, says its advertising harness earned the highest ad score in the TasteBench evaluation, reaching 8.8 and beating all six tested ChatGPT and Gemini configurations. The next-best result, ChatGPT using GPT-6 Astra at medium reasoning, scored 7.6.
- Omneky's advertising harness scored 8.8 on TasteBench, according to the company.
- The next-best configuration, ChatGPT with GPT-6 Astra at medium reasoning, scored 7.6.
- ChatGPT scores ranged from 7.3 to 7.6 and Gemini scores from 6.8 to 7.2.
- 60% of Omneky's ads passed every applicable check, versus 13% for the next-best configuration.
- The evaluation covered 52 ad requests across four brands, produced 364 ads, and does not measure campaign returns.
How the scores compare
According to Omneky, ChatGPT ad scores across the tested configurations ranged from 7.3 to 7.6, while Gemini scores ranged from 6.8 to 7.2. Omneky led the nearest competitor by 1.2 points.
The comparison covered 52 ad requests across four brands, which produced 364 ads for evaluation. Omneky conducted the evaluation itself, using three anonymous judges and applying the same standard to each tested system.
The company also reported results on a separate readiness measure. Sixty percent of Omneky's ads passed every applicable check, compared with 13% for the next-best configuration. According to Omneky, this supporting metric shows whether an ad meets all of the benchmark's required checks.
What a harness is and why it matters here
Omneky frames the result as evidence of the role the harness plays in creative production. It defines a harness as the software surrounding a model that supplies context and directs its work.
In Omneky's case, the advertising workflow reads brand information, develops art direction, provides product and logo references, and selects an image model. TasteBench compares the ads produced by these complete workflows rather than the underlying models alone.
The ChatGPT and Gemini runs used APIs with web search and image generation. Omneky says app spot-checks produced similar results.
Company statement and limits of the findings
"Creative is one of the most important levers for growth, and the quality of the finished ad matters to every marketer," said Hikari Senju, founder and CEO of Omneky. He added: "We built Omneky's harness to bring brand knowledge and creative direction into every generation."
The company notes that the findings assess generated creative and do not measure campaign returns. A high ad score therefore says nothing directly about how an ad performs once it runs in a campaign.
Because the evaluation was run by Omneky, the results are the company's own claims. Omneky has published the results and side-by-side examples at www.omneky.com/tastebench.
What it means for marketers
For marketers, the announcement points to a question beyond which model is used: how much brand context and creative direction surrounds it. Omneky describes itself as a platform that helps businesses create image and video ads, review campaign performance and launch campaigns, with an agent that brings these workflows into a single conversation.
Anyone weighing the results should keep in mind that they come from a vendor-run benchmark covering 52 requests and four brands, and that they measure the quality of generated ads, not business outcomes.
PR Newswire: Omneky Harness Outperforms ChatGPT and Gemini on TasteBench Ad Score


