Model Detail
ZAYA1-8B
—ZAYA1-8B is an AI model with 8B parameters released by Zyphra. Distributed under the permissive apache-2.0 license.
ZAYA1-8B ships with 8B parameters. Total weight footprint is approximately 8.8 GB, which is the relevant figure when planning local-inference VRAM. The apache-2.0 license is permissive, allowing commercial deployment and derivative work without per-seat fees, though attribution requirements still apply.
Downloads of ZAYA1-8B have moved +1973.5% over the trailing seven days. That puts the model in active uptrend territory; a sustained move of this size usually reflects a recent release, a viral integration, or a benchmark surprise rather than steady-state demand. These numbers are signal, not guarantee — week-over-week download counts on Hugging Face also reflect mirror traffic, CI scrapes, and one-off benchmarking runs.
ZAYA1-8B is best fit for general-purpose AI workloads. Treat this as a starting matrix rather than a benchmark verdict — the right deployment usually depends on the specific evaluation suite that mirrors your workload.
ZAYA1-8B Technical Report
arXiv:2605.05365v1 Announce Type: new Abstract: We present ZAYA1-8B, a reasoning-focused mixture-of-experts (MoE) model with 700M active and 8B total parameters, built on Zyphra's MoE++ architecture. ZAYA1-8B's core pretraining, midtraining, and supervised fine-tuning (SFT) were performed on a full-
ZAYA1-VL-8B Technical Report
arXiv:2605.08560v1 Announce Type: cross Abstract: We present ZAYA1-VL-8B, a compact mixture-of-experts vision-language model built upon our in-house language model, ZAYA1-8B. Despite its compact size, ZAYA1-VL achieves performance competitive with leading base models such as Molmo2-4B and InternVL3.
