MemTensor/MemT-1.0-Flash
archivedmodel
MemT-1.0 is a new generation of memory foundation models that makes persistent, dynamically evolving memory an intrinsic part of model computation. Mixture-of-experts, 288B total (256 experts, 6 per token). Context: 1M tokens (config). License: mit. Vendor-reported: "Across 11 memory benchmarks, MemT-1.0-Flash and MemT-1.0-Nano rank first and second among the evaluated parametric-memory methods, respectively."
- Vendor-reported: Across 11 memory benchmarks, MemT-1.0-Flash and MemT-1.0-Nano rank first and second among the evaluated parametric-memory methods, respectively.
- Org
- MemTensor
- Released
- 2026-10-17
- Status
- archived
- License
- mit (permissive)
- Parameters
- 287.9B
- Architecture
- MoE · 256 experts, 6/token · MemTDeepseekV4ForCausalLM
- Context
- 1,048,576 tokens
- Task
- text-generation · text
- Signal
- 70 HF likes · 682 downloads
- Archived
- 2026-10-22 (no longer listed by the collector; stats as last seen)
Runs on
estimate from params; need = weights + FP16 KV cache for 8K tokens + 1GB. A sizing calculation, not tested.
- Q4 ~174.5 GB (estimated size): 2× DGX Spark (256GB), Mac 256GB, CPU only, 256GB RAM (slow)
- Q8 ~307.6 GB (estimated size): Mac 512GB
- FP16 ~577.6 GB (estimated size): no listed device
Links
First published here 2026-10-17 · updated 2026-10-22 · markdown · JSON