Nemotron 3 Ultra is built for, frontier reasoning, orchestration, coding agents, deep research, and complex enterprise workflows. It delivers up to 5x faster inference and up to 30% lower cost for agentic workloads while supporting up to 1M token context.
Nemotron 3 Diarization is an open-weight streaming Sortformer model, engineered for low-latency speaker diarization across up to 8 speakers, ordered by arrival time. It powers meeting transcription, call analytics, and voice-agent pipelines—resolving who spoke when with generic speaker labels and timestamps in a single streaming pass, offline or in real time.
MiMo-V2.6-Pro is the flagship of Xiaomi's MiMo-V2.6 series: a sparse Mixture-of-Experts model with 1.02T total and 42B activated parameters, natively omnimodal across text, image, video and audio, with a 1M-token context window. Trained with large-scale mixed reinforcement learning and groupwise agentic grading, it is built for demanding agentic work including long-horizon coding, multi-tool agents and cybersecurity tasks.
MiMo-V2.6-Flash is the efficiency-balanced model in Xiaomi's MiMo-V2.6 series: a sparse Mixture-of-Experts model with 309B total and 15B activated parameters, natively omnimodal across text, image, video and audio, with a 1M-token context window. Trained with large-scale mixed reinforcement learning across coding, general agents, visual and cybersecurity tasks, it is built for agentic workloads such as long-horizon coding, tool use and computer use.
Qwen's fast, low-cost model with a one-million-token context window, billed at a flat rate across the entire window.
Qwen3.8 27B is an open-weight dense vision-language model from Qwen. It is suited for coding, professional workflows, research, multimodal interaction, and long-running agent tasks, with flexible thinking that can be enabled or disabled.
Open-weight layer decomposition (6B) from the Ming-Image 0.1 Design family — splits a flattened design or composite into stacked, transparent, editable PNG layers (text, subject, background) for localizable, editable design workflows. MIT-licensed.
Design-native, open-weight text-to-image (6B) — ranked #1 among open-weight models on Artificial Analysis's UI/UX Design leaderboard. Renders legible UI and poster text plus cohesive graphics, illustration and photography in ~1.7s. MIT-licensed.
Nemotron 3 Diarization is an open-weight streaming Sortformer model, engineered for low-latency speaker diarization across up to 8 speakers, ordered by arrival time. It powers meeting transcription, call analytics, and voice-agent pipelines—resolving who spoke when with generic speaker labels and timestamps in a single streaming pass, offline or in real time.
DeepSeek-V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism that reduces training and inference cost while preserving quality in long-context scenarios. A scalable reinforcement learning post-training framework further improves reasoning, with reported performance in the GPT-5 class, and the model has demonstrated gold-medal results on the 2025 IMO and IOI. V3.2 also uses a large-scale agentic task synthesis pipeline to better integrate reasoning into tool-use settings, boosting compliance and generalization in interactive environments.
Hy3 is a 295B-parameter Mixture-of-Experts (MoE) model with 21B active parameters and 3.8B MTP layer parameters, developed by the Tencent Hy Team. Following the Hy3 Preview launch in late April, we gathered feedback from 50+ products and scaled up post-training with higher quality data. Today, we introduce Hy3, which outperforms similar-size models and rivals flagship open-source models with 2-5x parameters. It also shows significant gains in utility across various products and productivity tasks.
Introduce DeepSeek-V4.1-Flash, a multimodal Mixture-of-Experts (MoE) model with 552B backbone parameters and support for contexts of up to one million tokens.
Realtime model with extended thinking for complex live reasoning
Realtime multimodal model for live audio/video interactions
Latest fast multimodal model with strong general capabilities
DeepSeek-V3-0324, a strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token, an improved iteration over DeepSeek-V3.