Baidu has launched ERNIE X1 Turbo and 4.5 Turbo, two high-speed models that boast impressive performance while significantly reducing costs.
As an enhancement to the existing ERNIE X1 and 4.5 models, these new Turbo versions emphasize multimodal processing, strong reasoning capabilities, and aggressive pricing strategies, aiming to attract developers' interest and gain market share.
Baidu ERNIE X1 Turbo: Combining Deep Reasoning with Cost Efficiency
ERNIE X1 Turbo is positioned as a deep reasoning model capable of handling complex tasks that require advanced comprehension. It enters a highly competitive field, claiming superior performance in certain benchmarks compared to rivals like DeepSeek R1, V3, and OpenAI o1:
A key feature of X1 Turbo is its advanced "chain-of-thought" process, which allows for more structured and logical problem-solving.
In addition, ERNIE X1 Turbo has enhanced multimodal capabilities—able to understand and process information beyond text, possibly including images or other data types—and improved tool utilization. This makes it particularly suitable for nuanced applications such as literary creation, complex logical reasoning challenges, code generation, and intricate instruction following.
ERNIE X1 Turbo achieves this performance at a lower price point than its competitors. Input token costs start at $0.14 per million tokens, with output tokens priced at $0.55 per million. This pricing structure is approximately 25% of DeepSeek R1's cost.
Baidu ERNIE 4.5 Turbo: Multimodal Power at a Fraction of the Cost
Sharing the spotlight is ERNIE 4.5 Turbo, which focuses on delivering upgraded multimodal capabilities and significantly faster response times compared to non-Turbo counterparts. The emphasis here is on providing a versatile, responsive AI experience while substantially lowering operational costs.
Compared to the original ERNIE 4.5, this model achieves an 80% price reduction, with input costs set at $0.11 per million tokens and output at $0.44 per million. This is about 40% of the latest version of DeepSeek V3’s cost, further highlighting a well-thought-out strategy to attract users through cost efficiency.
Performance benchmarks further solidify its credentials. In multiple tests evaluating multimodal and text functions, Baidu ERNIE 4.5 Turbo outperforms OpenAI's acclaimed GPT-4o model.
In multimodal capability assessments, ERNIE 4.5 Turbo scores an average of 77.68, surpassing GPT-4o’s score of 72.76 in the same test.
While benchmark results always need careful interpretation, this indicates that ERNIE 4.5 Turbo is a formidable contender for tasks involving comprehensive understanding of different data types.
Baidu Continues to Shake Up the AI Market
The launch of ERNIE X1 Turbo and 4.5 Turbo marks a growing trend in the AI domain: the democratization of high-end functionalities. While foundational models continually push the boundaries of performance, there's an increasing demand for models that balance functionality, accessibility, and economy.
By lowering the price point of models with complex reasoning and multimodal features, the Baidu ERNIE Turbo series could enable a broader range of developers and enterprises to integrate advanced AI into their applications.
This competitive pricing puts pressure on established players like OpenAI and Anthropic, as well as emerging competitors like DeepSeek, potentially leading to further price adjustments across the market.
