Interfaze releases Lite weights for document and speech tasks

The download combines a reasoning core with specialist tools and needs a GPU with 80 GB of memory.

Maya Chen
October 5, 2026
AI Model Releases, Models
1 min

Black and white microphone on a table in an illustrative file photograph — Illustrative microphone file photograph unrelated to Interfaze hardware

Listen to this article

Interfaze published the weights for Interfaze 1 Lite on October 5, with the release recorded in its Hugging Face activity feed. The company describes a reasoning core working with specialist models for document reading, speech transcription and visual tasks.

Its announcement says the model runs on one GPU with 80 GB of memory and can return text or structured data. The repository is labeled Apache 2.0.

Local deployment has limits

The model card warns that dense document pages can take a minute or more through Transformers, and large PDFs can raise memory use. Local chat does not accept a response schema, while the hosted API supports one. Benchmark results are reported by Interfaze.

Related coverage looks at Cactus Whistle for local speech recognition.

Illustrative microphone file photograph by Dušan Cvetanović on Pexels, published June 18, 2022 under the Pexels License. The equipment shown is unrelated to Interfaze.

Maya Chen

Maya Chen is focused on covering AI models, research, and the evidence behind new capabilities. Maya follows model launches, benchmarks, open weights, and scientific uses of AI with one question in mind. What changed, and how would we know? The voice is curious and exacting, with a soft spot for elegant technical ideas and little patience for a leaderboard without context.