eWEEK content material and product suggestions are editorially unbiased. We might become profitable once you click on on hyperlinks to our companions. Learn More.
Headquartered in Shenzhen, China, the staff with Tencent lately unveiled their new AI platform referred to as Hunyuan Turbo S. Designed particularly as a competitor to DeepSeek, which was additionally created by a Chinese AI firm, Tencent hopes its generative AI platform will assist it acquire recognition amongst the highest AI firms on the planet.
Hunyuan Turbo S is able to replying to consumer inputs and queries inside one second, which is even quicker than DeepSeek-R1, in response to the corporate and as reported by Reuters. We haven’t discovered any pace benchmarks that affirm Tencent’s declare.
How Hunyuan Turbo S compares to opponents within the benchmarks
According to benchmarks supplied by Tencent as reported by WinBuzzer, Hunyuan Turbo S leads many opponents in a wide range of areas. The following benchmarks are generally used to guage the performance, effectivity, and accuracy of huge language fashions (LLMs).
- Chinese: Hunyuan Turbo S ranks the best in Chinese language benchmarks carried out by CMMLU, however DeepSeek-R1-Zero leads in C-Eval’s benchmarks.
- Alignment: Although Hunyuan Turbo S outperforms GPT-4o, Claude 3.5, Llama 3.1, and DeepSeek-V3 in benchmarks from LiveBench, it lags barely behind Claude 3.5 in benchmarks from IF-Eval.
Some of Hunyuan Turbo S’s weaknesses embody:
- Math: Hunyuan Turbo S outperforms GPT-4o, Claude 3.5, Llama 3.1, and DeepSeek-V3 in some benchmarks, however DeepSeek-R1-Zero leads all of them as scored by AIME 2024 and MATH.
- Knowledge: Hunyuan Turbo S ranks pretty excessive on most data benchmarks, however it doesn’t fairly match as much as DeepSeek-R1-Zero within the benchmarks from MMLU, MMLU-Pro, and SimpleQA.
- Reasoning: Hunyuan Turbo S solely ranks third-highest, behind GPT-4o and Claude 3.5, on BBH’s reasoning benchmarks.
- Code: While HumanEval has Hunyuan Turbo S sitting proper behind Claude for coding capabilities, it’s a bit additional behind DeepSeek-V3, DeepSeek-R1-Zero, and GPT-4o on StayCodeBench’s outcomes.
While Hunyuan Turbo S is the clear winner in sure circumstances, it nonetheless falls behind DeepSeek-R1-Zero in a number of situations.
A powerful competitor within the AI race
Tencent’s new Hunyuan Turbo S platform solidifies the Chinese tech big’s place within the race to develop the quickest and strongest AI platform. Although it’s not Tencent’s first foray into the world of generative AI instruments, it’s the firm’s most noteworthy entry up to now – and it’s definitely one to look at over the approaching weeks, months, and years.
