NVIDIA

NVIDIA has published Qwen3.8-Flash-Next-FP4-Local, a 96 GiB NVFP4 quantization of Alibaba's 125B-parameter MoE model, on Hugging Face. NVIDIA reports 91.5% on GPQA Diamond versus 92.2% for BF16. It requires NVIDIA's custom vLLM containers because upstream support has not landed.

NVIDIA and IIT Madras' AI4Bharat lab announced Nemotron-pretraining-Indic-v1, a corpus of roughly 2.8 trillion tokens covering all 22 languages in India's Constitution. It is due for release soon, alongside the open GPU-accelerated pipeline used to generate it.

Ollama v0.40.5 adds NVIDIA Nemotron 3.5 Super, a 120B-parameter hybrid Mamba-2/Transformer MoE vision-language model with 12B active parameters and context up to 1M tokens. It also adds a smaller sibling, Nemotron 3.5 Bolt, and fixes a Responses API bug with custom tool calls.

Nvidia has released Nemotron 3.5 Super VL, an open multimodal Mixture-of-Experts model with 120 billion total and 12 billion active parameters, built for continuously running enterprise agents. Two compact Bolt models for local PCs and edge devices launch alongside it, though the 4 GB variant ships later this month.

NVIDIA reports that specialized versions of Nemotron 3 Ultra reached gold-medal level at both IOI 2026 (535.4/600) and IMO 2026 (30/42). The IOI run was unofficial, and the IMO proofs were graded by official IMO graders, according to NVIDIA. Checkpoints, datasets, and inference pipelines are published on Hugging Face.

A new Nvidia research system called SoL-Pi automatically rewrites the control logic of coding agents rather than the underlying model, cutting token usage by up to 49% while keeping performance nearly intact. The approach could shift efficiency gains in AI agents from model-level tricks to harness-level engineering.

NVIDIA released Nemotron 3 Diarization, a 100-million-parameter open-weight model that identifies who is speaking and when in audio conversations. It ranked #1 among 17 system configurations on Voice Arena's Diarization-Bench with a 14.72% diarization error rate, supporting up to eight speakers in both live and recorded audio.

Nvidia and Palantir are integrating Nvidia's open Nemotron models and cuOpt optimization engine into Palantir's Foundry platform to manage complex supply chains. The first deployment is Nvidia's own operation, where a single Vera Rubin server rack contains 1.3 million components.

Nvidia's new Groq 3 LPX inference accelerator hit 3,400 tokens per second on Gemma 4 31B, a figure the company says is four times faster than Cerebras. Experts note the benchmark uses at least 64 LPX chips versus Cerebras' one or two accelerators, making the comparison far less clean than it appears.

Nvidia researchers report that a custom agent harness with a supervisory component drove Claude Opus 5 from a 30% baseline score to a perfect 100% on the ARC-AGI-3 benchmark. The findings add to mounting evidence that scaffolding, not the underlying model, is the dominant factor in long-horizon agentic performance.

NVIDIA's Nemotron 3.5 Lightning, a 30B-parameter hybrid Mixture-of-Experts model with only 3B active parameters, is now available for one-click deployment on Amazon SageMaker JumpStart. NVIDIA claims up to 4x higher throughput and 30% faster task completion for high-volume agentic workloads compared to larger frontier models.

NVIDIA released the full-precision BF16 reference weights for Nemotron 3.5 Lightning, a 30B-parameter Mixture-of-Experts model with only 3B active parameters and support for up to 1 million tokens of context. The model uses a hybrid Mamba-2, MoE, and Attention architecture and is licensed under OpenMDW-1.1 for commercial use.

Nvidia is reportedly building Nemotron 4, an open-weight model with at least one trillion parameters — double the size of Nemotron 3 Ultra. The company has tripled its cloud spending on in-house training to $28 billion through 2031, with an earliest possible release this fall.

NVIDIA has published NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4, a 30-billion-parameter Mixture-of-Experts model with only 3B active parameters, a hybrid Mamba-2/MoE/Attention architecture, and support for up to 1 million tokens of context. The NVFP4-quantized checkpoint is designed to run on a single DGX Spark (GB10) or H100 GPU.

Nvidia's Nemotron 3.5 Lightning, a 31.6B-parameter open-weight model with only 3.6B active parameters, matches OpenAI's gpt-oss-120b on the Artificial Analysis Intelligence Index while delivering the fastest throughput in its class at nearly 670 tokens per second. The model posts especially large gains on agentic benchmarks, beating both gpt-oss-120b and the larger Nemotron 3 Super.

NVIDIA has released Alpamayo 2 Super, a 34B-parameter foundation model for autonomous vehicle development that combines a 32B vision-language backbone with a 2.3B-parameter diffusion action decoder. The model handles trajectory prediction, visual question answering, 2D grounding, and auto-labeling, and posts a Lingo-Judge score of 79.2 on LingoQA reasoning evaluation.

NVIDIA's Magpie TTS Multilingual, a 364M-parameter open-weights text-to-speech model, now supports 12 languages after adding Modern Standard Arabic, Korean, and Brazilian Portuguese. The model achieves 32ms time-to-first-audio on B200 GPUs and improves speech quality across French, Spanish, and German.

NVIDIA has released NemotronLabs VoiceChat 11B, an 11-billion-parameter end-to-end full-duplex speech model that unifies streaming speech understanding and generation in one architecture. The model claims to be the first open full-duplex system to support live tool calling during natural conversation, with ~450ms turn-taking latency.

NVIDIA has released Cosmos 3 Edge, a 4-billion-parameter open world model designed for edge AI systems. The model delivers real-time robot control at 15 Hz on NVIDIA Jetson devices, generating 32 actions per inference at 640×360 resolution.

NVIDIA has released Nemotron-3-Embed-1B-BF16, a 1.14 billion parameter text embedding model supporting 34 languages with a 32,768 token context window. The model generates 2048-dimensional embeddings and was derived from Ministral-3-3B-Instruct-2512 through two rounds of structured pruning and distillation, first to 2B then to 1.14B parameters.