TML Interaction
#11 in Reasoning Modelsthinking-machines · seit 2026-05-11 · 19× · zuletzt 30. Juni 2026
39
Momentum
TML-Interaction-Small is the first model from Thinking Machines Lab (founded by Mira Murati, former CTO of OpenAI). It is a 276-billion-parameter Mixture-of-Experts (MoE) model with 12 billion active parameters, released as a research preview on May 11, 2026. The model processes audio, video, and text as simultaneous, continuous streams in 200-millisecond micro-turn chunks (full-duplex), without external voice-activity detection or turn-boundary systems. It consists of two components: a live interaction model for real-time conversation and an asynchronous background model for complex reasoning and tool-use tasks.
Momentum trend
04.04.03.07.
Features
| Key Benchmark (%) | FD-bench v1.5 (interaction quality): 77.8 vs. 54.3 (Gemini-3.1-flash-live minimal) and 46.8 (GPT-realtime-2.0 minimal) |
| License | No open access/no open weights; gated Research Preview, access only upon request via interaction@thinkingmachines.ai |
| Multimodality | Input: continuous audio, video/images, text; Output: audio and text – all modalities processed simultaneously in 200ms micro-turns, encoder-free early-fusion architecture |
| Platform | Cloud-based Research Preview (select partners only); serving optimizations upstreamed to, among others, the SGLang inference framework |
| Price per 1M Tokens | Not yet published; pricing to be announced only at wider release |
| Release Date | May 11, 2026 (blog post 'Interaction Models: A Scalable Approach to Human-AI Collaboration'); limited Research Preview to follow in coming months, wider release later in 2026 |
