Mk1.5 Real-Time Multimodal Reasoning Model Launches with 2-5x Lower Latency

Mk1.5, a multimodal model for real-world real-time applications, launches with 2-5x lower latency than its predecessor, adding video localization and audio-visual understanding, and relying on externalized knowledge via tool calls and parallel sub-agents.

2026-09-25 ~ 2026-09-25 · 4 related posts

• Mk1.5 real-time multimodal reasoning model ships with 2-5x lower latency than Mk1 — AkshatS07 · 2026-09-25
• Mk1.5 Adds Video Localization, Audio and Audio-Visual Understanding — AkshatS07 · 2026-09-25
• Mk1.5 explained: factor world knowledge into tool calls, parallel research via sub-agents — AkshatS07 · 2026-09-25
• Mk1.5 adds video localization, expands into audio and audio-visual understanding — AkshatS07 · 2026-09-25