AI models
Tiny Recursive Model (TRM-Att)
Training compute
3.1×10²⁰ FLOP
Parameters
7M
Published
Oct 6, 2025
Tiny Recursive Model (TRM-Att) is an AI model developed by Samsung SAIT AI Lab (South Korea), first published in October 2025. It works in the language, vision and multimodal domain, on tasks such as language modeling/generation, question answering and visual puzzles.
Training it took an estimated 3.1×10²⁰ FLOP of compute (estimation method: hardware). The model has 7,000,000 parameters. Training ran on 4 NVIDIA H100 SXM5 80GB for about 72 hours.
Access: Unreleased. Its weights are not openly released. Epoch AI rates the confidence of this record as confident.
Full record
- Organization
- Samsung SAIT AI Lab
- Country of organization
- South Korea
- Domain
- Language, Vision, Multimodal
- Task
- Language modeling/generation, Question answering, Visual puzzles
- Training compute
- 3.1×10²⁰ FLOP
- Compute estimation method
- Hardware
- Parameters
- 7,000,000
- Training hardware
- NVIDIA H100 SXM5 80GB
- Chips used
- 4
- Training time
- 72 h
- Training power draw
- 5.5 kW
- Model accessibility
- Unreleased
- Open weights
- No
- Epoch confidence
- Confident
SourceEpoch AI, 'AI Models'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/ai-models. Licensed under CC BY 4.0.
← All ai models