Handles text and speech input, producing text, speech, or both for streaming voice interaction, multilingual audio tasks, and spoken tool calls. Its 2B-parameter size targets local and edge use, not complex multi-step reasoning.
Provides low-latency, true-streaming automatic speech recognition that emits append-only committed transcripts to avoid partial-result rollbacks. Supports configurable decoding chunks (80 ms–2 s), optimized for Chinese and English, and offers vLLM and transformers backends for real-time deployment.
Preview agentic language model for research and engineering workflows that turns research questions into executable, verifiable workflows via tool use and long-context reasoning; built on a 744B-parameter MoE (GLM-5.2) with MIT-licensed BF16 and FP8 checkpoints.
A pretrained autoregressive foundation LLM from Yandex that uses a hybrid MoE design (80B params, 3B active) and supports up to 262,144-token context. Notably strong on Russian factual knowledge, math and coding benchmarks; intended for research and fine-tuning rather than out-of-the-box production.
Generates unified 768‑dimensional embeddings for text (including code), images, video and audio to enable cross‑modal semantic search and retrieval. Supports task instruction prefixes, Matryoshka truncation to 128/256/512/768 dims, and modular encoders for on‑device use under an Apache‑2.0 license.
Provides transcripts, translations, and multi-task text labels for speech across 48 languages and about 463,000 hours. The release contains manifests only; audio must be obtained from the original data providers.
A non-autoregressive, multilingual System‑1 decision model that answers typed questions (choice/score/noul) over text or JSON in a single forward pass with probability distributions. Key differences: calibrated-probability design, sub-35ms GPU latency, and built-in multilingual routing; requires per-workflow recalibration and validation on long or low-resource languages.
Maps a state and question to typed probabilistic decisions (choice distributions, yes/no probabilities, or scored/ordinal outputs) across controlled synthetic tasks. Offers multiple frozen configs with train/calibration/validation/test/OOD splits, Parquet + raw JSONL exports, and reproducible manifests and provenance.
An all-in-one multilingual scene text recognition approach that pairs a shared visual encoder with a script-aware Mixture-of-Experts (ScriptMoE) decoder to route each image to top-2 script experts. Introduces TextMuSS-10M, a synthetic dataset covering 10 scripts and 229 languages, and reports state-of-the-art accuracy and large end-to-end OCR F1 gains while remaining parameter-efficient.
Turns a context, a question, and 2–20 candidate answers into a single chosen option for classification, routing, ordered scores, and Boolean decisions. Built on mmBERT-small with a 144.3M-parameter decision head, runs on CPU with up to 8,192 combined tokens; FP32 weights occupy 550.5 MiB and are Apache‑2.0 licensed.
Provides 98,877 historical newspaper page images (1700s–1940s) paired with ALTO OCR, per-word confidences and line/word bounding boxes in image pixels — ready for line-level OCR training, OCR quality estimation, re‑OCR comparisons and layout analysis. OCR is library-produced (silver); image resolutions and OCR quality vary.
Processes text and images, with support for reasoning, coding, and tool use. Its trillion-parameter MoE activates about 49B parameters per token and supports a 1M-token context.