Search “Hugging Face”

Mistral Releases Mistral Large 4 Weights: Coding Ability Leads Open-Source Models but Overall Trails Chinese Models

Mistral has released the weights of Mistral Large 4 on Hugging Face. The model, nicknamed Le Chonk, has approximately 1 trillion parameters with 52 billion activated per step, and supports text and image inputs. On Terminal-Bench 4.0, its score rose from 28% in the preview version to 41%, slightly above GLM-5.3's 40%, but on the new version of the Artificial Analysis Intelligence Index it scored only 26 points, below the 32 points of DeepSeek V4.1 Flash and GLM-5.3. API pricing is $0.70 per million input tokens and $2.10 per million output tokens, under a modified MIT license; using it as a commercial cloud service requires separate authorization.

NVIDIA Releases Nemotron 3.5 Transcribe and Granary v2 Data Pipeline

NVIDIA has released the open-weight model Nemotron 3.5 Transcribe on Hugging Face and ModelScope, integrating transcription, speaker diarization, translation, and audio question answering into a single checkpoint. It supports transcription in 40 languages, translation between 31 languages and English, and speaker labeling in 13 languages, can process recordings up to 1 hour long, and can label up to 8 speakers. The model ranks first among open-weight models on the English leaderboard of the Hugging Face Open ASR Leaderboard with an average WER of 5.33%, and achieves 2.8% AA-WER on Artificial Analysis with approximately 4B activated parameters. The simultaneously released Granary v2 covers 48 languages and 463,062 hours of speech data, and the accompanying NeMo Speech fine-tuning recipe reduced WER on held-out speakers from 22.44% to 12.17% in a medical case.

Nvidia Releases Open Source Transcription Model Nemotron 3.5 Transcribe

Nvidia has released the free open-weight model Nemotron 3.5 Transcribe, which can transcribe audio in up to 40 languages, and simultaneously launched version 2 of the open speech dataset Granary. The model supports translation between 31 languages and English, processes recordings up to one hour long, and distinguishes up to 8 speakers; Nvidia claims it outperforms competing products on benchmarks such as the Hugging Face Open ASR Leaderboard. The model has 30 billion parameters and can run on Jetson AGX Thor, DGX Spark, and Blackwell data center chips, but it only supports recorded audio and does not support live streams.

SoMark Open-Sources 1.3B Document Parsing Model OXR-1.0

SoMark has open-sourced OXR-1.0, a 1.3B-parameter document parsing model that converts PDFs and page images into Markdown or JSON. The weights have been released on Hugging Face, and lightweight testing can run on a single 24GB GPU. The model ranks first on olmOCR-Bench, OmniDocBench, and DocParsingBench, achieves a decoding speed of 558 tokens/s on a single NVIDIA RTX PRO 6000 GPU, and parses 100 pages end-to-end in about 101 seconds. Its post-training stage employs a recursive self-improvement (RSI) loop, in which a general-purpose VLM judges parsing results and generates newly annotated pages, eliminating the need for manual annotation.

Ant's Ling 3.1 Flash 1019 Open-Sourced, Intelligence Index Rises to 44

Ant Group's AntLingAGI has released Ling 3.1 Flash 1019, an updated version of Ling 3.1 Flash, scoring 44 on the Artificial Analysis Intelligence Index v4.3, up from 41 for the previous generation. The model is a reasoning model with 560B total parameters and 25B activated parameters, supports a 1M token context window, is released under the MIT license, is available as open weights on Hugging Face, and can be accessed via InclusionAI's first-party API. In terms of agentic capabilities, AA-Briefcase v1.1's Elo rose from 1397 to 1521, GDPval-AA v2.1 rose from 1617 to 1653, and AutomationBench-AA rose from 62% to 68%; Humanity's Last Exam rose from 39% to 43%, and CritPt rose from 18% to 26%. Output tokens per task increased to 145k, up from 103k for the previous generation, with a cost of $1.49 per task and pricing of $0.30 per 1M input tokens and $0.90 per 1M output tokens.

Liquid AI Releases Six 2B-Class Open-Weight Models in the LFM3 Series

On October 27, Liquid AI released the LFM3 series. All six 2B-parameter-class open-weight models branch from the same LFM3-1.7B-Base, covering text, vision, audio, omni-modal, and decision models. LFM3-1.7B has 1.73B parameters, a 256K token context, supports 16 languages, and adopts a Tri-Hybrid architecture with 13 layers of gated short convolutions, 5 layers of sliding-window attention, and 6 layers of global attention. It is available under the LFM Open License v1.0 on Hugging Face, the Liquid AI Playground, and via API. The company claims that LFM3-1.7B surpasses LFM2.5-2.6B with 35% fewer parameters and occupies about 1 GB of memory at 4-bit precision, but the model card states that it is not suitable for agentic coding or complex multi-step reasoning.

Liquid AI Releases the LFM3 Series of Models, Including 1.7B Text, Vision, Audio, and Omni Versions

Liquid AI has released the LFM3 series of open-weight models, all branched from LFM3-1.7B-Base, covering text, vision, audio, Omni, and the d1 decision model. LFM3-1.7B-Base adopts a three-hybrid-layer architecture, is pretrained on 22T tokens, and in the mid-training stage over the final 3.5T tokens incorporates more agent trajectories and tool-calling data, with the context window extended to 256K tokens. The company says LFM3-1.7B is about 25% faster than LFM2-1.2B at an 8K context, and scores 38.42 versus 3.02 on WebArena-Verified; LFM3-VL-2B's average score across 62 vision benchmarks rises from 57.6 to 71.4. All models are available for download on Hugging Face, along with a single-GPU fine-tuning notebook.

Baseten releases decision model joey-1, with P50 latency of 20ms and $0.01 per million tokens

Baseten's Base Labs has released joey-1, a 2B decision model that directly outputs a single-token label and calibrated confidence, with a P50 end-to-end latency of 20ms, about 2.5 times faster than elephant-1 (49ms), at a price of $0.01 per million tokens, one-tenth that of mammoth-1. joey-1 is post-trained from Qwen3.5-2B and shares the same data recipe and API with elephant-1 and mammoth-1, so switching requires changing only one line. The model is now available on Baseten Model APIs, with zero data retention by default, and its weights are open-sourced on Hugging Face; Model APIs supports a 16k context window, while dedicated deployments can unlock a 262k context window and support fine-tuned weights.

Mistral Releases Weights for Trillion-Parameter Open Model Mistral Large 4, Focused on Cybersecurity

Mistral has released the weights for its trillion-parameter open-source model Mistral Large 4 (ML4, nicknamed Le Chonk), which enterprises and government agencies can download and deploy themselves; the weights are already available on Hugging Face and via its API. Mistral says the model's performance is comparable to that of the open-source model GLM-5.3, but it consumes 47% fewer tokens per task, and it ranks first among open-source models on Artificial Analysis's Coding Agent Index. Pierre Stock, Mistral's Vice President of Science, said that open weights mean organizations can deploy locally in their own environments to identify, coordinate, and respond to major security incidents in real time.

Baseten Releases 2B Decision Model joey-1, Priced at $0.01 per Million Tokens

Baseten's Base Labs has released a 2B decision model, joey-1, post-trained on Qwen3.5-2B, which directly returns labels and calibrated probabilities instead of generating text, targeting high-concurrency routing, classification, and guardrail checks. The model scores 63.30 on the public Jev Decision Index 0.4 set, the highest among models under 4B, but 8.94 points lower than the 9B elephant-1. Model APIs are priced at $0.01 per million input tokens, with a P50 latency of 20 ms for an average 2000-token request; the weights are open-sourced under the MIT license on Hugging Face and can be self-hosted on a single CUDA GPU.

Mistral Releases Flagship Model Mistral Large 4 and Opens Weight Downloads

Mistral AI has released its new flagship frontier model, Mistral Large 4, codenamed Le Chonk, making it available for download after ending its public preview. The model supports text and image processing, with a focus on strengthening defensive cybersecurity, agentic capabilities, and geospatial and grounding capabilities, and is said to achieve state-of-the-art results on agentic benchmarks such as SWE-bench and Terminal Bench. The model was trained on more than 160 languages and is available via Mistral Studio, the Mistral API, OpenRouter, Hugging Face open weights, Vibe Code, and Nvidia-hosted endpoints.

Tencent Hunyuan open-sources interactive video world model Hy-Nomas

Tencent Hunyuan releases the open-source free-form interactive video world model Hy-Nomas, which supports interaction using text prompts and camera trajectories, and features long-horizon generation with world memory. The model can be adapted to specialized domains such as robotics via lightweight LoRA, and has been optimized for consumer-grade GPU inference. The official release also provides website, GitHub, Hugging Face, and demo entry points.

MiniMax open-sources the M3.1 Flash model, reducing usage cost by 70%

MiniMax announced on October 23 the open-sourcing of the M3.1 Flash model, which processed over 100 trillion tokens within 72 hours of launch. Compared with the previous generation M3, M3.1 Flash has comprehensively improved capabilities, with usage cost reduced by 70% and online inference speed increased by nearly 3 times. MiniMax stated it will work with chip manufacturers such as Huawei Ascend, NVIDIA, and AMD, as well as inference frameworks and communities such as vLLM, SGLang, and Hugging Face, to advance day-one open-source integration.

MiniMax M3.1 Flash releases open-source weights, 428B total parameters, 23B activated per token

MiniMax M3.1 Flash weights are now public, with 428B total parameters and 23B activated per token. The company claims substantial improvements in agentic tasks, coding, and cybersecurity: BankerToolBench 76.2, DeepSWE v1.1 73.7%, OSWorld V2 61.7, Agents' Last Exam 29.6, CyberGym 82.2. The model uses MiniMax Sparse Attention and DSpark speculative decoding to improve inference efficiency; the weights are now available on Hugging Face and ModelScope, with API and MCode desktop entry points also provided.

Recursive releases Flint, a 40B MoE coding model, at 125+ tokens/s on a MacBook Pro

Recursive has released its first model, Flint, a 40B mixture-of-experts coding model under the Apache 2.0 license, with weights now open on Hugging Face, built from scratch in under 4 months by the automated research system Eureka Machine. Flint generates code at 125+ tokens/s on a MacBook Pro and reaches 150+ tokens/s on an RTX Spark. On a 48GB MacBook, its average quality is about 81%, compared with about 85% for Qwen3.8 27B, but in 128K-token long sessions it is about 7 times faster than the latter.

Recursive releases laptop-class coding model Flint, says it was built mainly with AI involvement

AI startup Recursive has released its first model, Flint, a coding model that runs offline on a laptop and is freely downloadable on Hugging Face under the Apache 2.0 license; the company says its development process was mainly carried out with the involvement of its in-house Eureka Machine. Flint is an MoE model that activates only 2.5B of its 40B parameters at a time, generates at about 131 tokens/sec on a MacBook Pro with an M4 Pro, and outperforms other models under 50B parameters on SWE-Bench Verified. Recursive says Flint requires more guidance than large paid coding assistants and recommends breaking large tasks into small steps.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

NVIDIA Releases Nemotron 3.5 Series of Open Models, Covering Embedding, Reranking, and Document Parsing

On October 21, 2026, NVIDIA expanded the Nemotron Retriever series, launching Nemotron 3.5 Embed (8B BF16, 1B BF16, 1B NVFP4), Nemotron 3.5 Rerank 8B BF16, and the 631-million-parameter Nemotron 3.5 Parse, aimed at multimodal retrieval and document processing for enterprise agent workflows. NVIDIA says that adding Rerank to the agent retrieval pipeline reduces the token cost per query by about 14% (with the 8B embedding model) and 13% (with the 1B model), respectively, while also improving retrieval accuracy. Parse scores 75.20 on ParseBench and processes 10.5 pages per second on a single H100; the open weights, data, and training recipes have been released on Hugging Face and can be deployed via NVIDIA NIM as well as Baseten, DeepInfra, and FriendliAI.