AI BriefingLiquid AI announces d1 with vision, a decision model matching GPT-6.1 Sol at lower cost
Key point
The new vision-capable decision model costs 19x to 200x less than GPT-6.1 Sol and Claude Opus 5.5 while delivering text decisions in 200 to 300 ms.
Details
Liquid AI has announced d1 with vision, an update to its first decision model that now accepts images, text, or both as inputs. The model is designed for classification and scoring tasks rather than generative output, providing probabilities for yes/no, choice, or score questions in a single forward pass without generating tokens.
Performance and Cost
In tests against GPT-6.1 Sol and Claude Opus 5.5 across six real-world applications, including filtering support tickets and inspecting circuit boards, d1 matched or outperformed GPT-6.1 Sol in four scenarios. The model offers significant efficiency gains, costing 19x to 200x less than both competitor models and answering significantly faster on every task tested.
Technical Details
The model achieves text decision times of 200 to 300 ms by avoiding token generation. It is available via the Liquid API at console.liquid.ai.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
SemiAnalysis: Anthropic Subscriptions Offer 5x More Value Than OpenAI
AWS Adds OpenAI GPT-6.1 Sol and Claude Sonnet 5.5 to Bedrock, Launches Managed Agents Preview