Model Detail
Qwen: Qwen3.8 Max Prime
—Qwen: Qwen3.8 Max Prime is a multimodal model released by Qwen. And supports text+image+video->text inputs.
Qwen: Qwen3.8 Max Prime is priced at $4/M input tokens and $12/M output tokens. Operationally the model offers a 1000K-token context window, which matters when sizing it for prompt-heavy or latency-sensitive workloads. This is frontier pricing, so cost-fit usually depends on whether the per-call quality justifies running a smaller commodity model two or three times to compare.
Qwen: Qwen3.8 Max Prime is published on Hugging Face but our pipeline has not yet captured architecture, license, or parameter-count metadata for this entry. The data is refreshed daily, so these fields typically populate within 24–48 hours of release.
Qwen: Qwen3.8 Max Prime is best fit for mixed text-and-image reasoning tasks such as document understanding, and long-context tasks such as full-codebase analysis or book-length summarization (1000K tokens). It is a less obvious choice for cost-sensitive batch processing where a commodity-tier model could be retried cheaply. 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.
Privacy-preserving inference on hosted models: a learned input-embedding transform for frozen Qwen3.8-27B
arXiv:2610.27817v1 Announce Type: new Abstract: Every prompt sent to a hosted language model is readable by whoever runs it. We study whether the prompt can remain on the client while the host receives transformed vectors that stand in for it. A language model converts text to vectors exactly once,
Qwen3.8-Omni: Towards Native Omni-Modal Agents
arXiv:2609.25611v1 Announce Type: new Abstract: We introduce Qwen3.8-Omni-Flash, a natively multimodal agentic model for real-world multimodal productivity. Compared with previous omni models, which primarily emphasized perception and interaction, Qwen3.8-Omni-Flash substantially improves multimodal
On the Design of Qwen3.8-Next Architecture: Evaluation, Efficiency, and Training Stability
arXiv:2608.30320v1 Announce Type: new Abstract: We describe the architecture and ablations of Qwen3.8-Flash-Next, a sparse mixture-of-experts model with 125B parameters, 6B activated per token, and additional 51B parameters of n-gram embedding tables held off the accelerator. On fourteen pre-trainin
