Thinking Machines Lab has launched Inkling, a new multimodal model designed for efficient reasoning and versatile task handling. The model accepts text, image, and audio inputs, and features controllable inference effort to balance reasoning depth with latency and token usage. Inkling incorporates several architectural innovations, including a mixture-of-experts approach and query-conditioned relative attention, and is available on Together AI's inference platform. AI
IMPACT Inkling's multimodal capabilities and controllable reasoning could enhance applications requiring nuanced understanding and adaptable performance.
RANK_REASON New model release from a lab, with details on architecture and capabilities. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
