NAll news

All news

Friday77 items

  1. QbitAIModel releases
    AI score81

    Google Releases Gemini 4 Carbon, Surpassing Claude Opus 5.5 on Multiple Benchmarks

    Google has released its new flagship model Gemini 4 Carbon, which tops the Artificial Analysis Intelligence Index 5.0 with a score of 59, surpassing Claude Opus 5.5 and GPT-6 Astra. It scored 65.2% on Terminal-Bench 4.0, 96.2% on GraphWalks for 256k to 1M context, and 88.2% on SEC-bench Pro vulnerability discovery. Initial pricing is $2 per million tokens for input and $10 per million tokens for output, with cached input as low as $0.2; because of its vulnerability discovery and exploitation capabilities, it will first be made available to select cybersecurity defenders.

    Same story · 34 reports

    #Model releases#Google#DeepMind

  2. QbitAIModel releases
    AI score72

    Diffusion Intelligence releases Clover-35B-A3B-Preview, a 35-billion-parameter diffusion language model

    Diffusion Intelligence releases the diffusion language model Clover-35B-A3B-Preview, which uses a MoE architecture with 35 billion total parameters and about 3 billion activated per token, and supports a thinking mode. On the 49 tasks of the web Agent benchmark WindTunnel, its median execution time was 1.52 seconds, lower than the 3.21 seconds of the Jev + Mercury 2.5 combination, with success rates of 98.6% and 95.9% respectively. On code benchmarks, it scored 73.6 on SWE-bench Verified and 51.7 on Terminal-Bench 2.1, and the team said it will subsequently launch a multimodal version and advance on-device adaptation.

    Same story · 2 reports

    #Model releases#Agents#English text is being prepared.

  3. The PaperModel releases
    AI score77

    Google Releases Gemini 4 Carbon, Strengthening Cybersecurity and Long-Running Agent Workflows

    Two weeks after the debut of Gemini 4 Argon, Google released a new-generation frontier model, Gemini 4 Carbon, saying it achieves the current highest level in cybersecurity and long-horizon tasks, with enhancements for coding, Agent workflows, and cyber defense. The model scored 88.2% on the vulnerability discovery benchmark SEC-bench Pro, 54.8% on the exploit benchmark Exploit Gym, and 71.0% on the patch application benchmark CWE-bench. Carbon will first be made available this week to some cybersecurity defenders, then expanded to more trusted partners, and only afterward released to developers, enterprises, and consumers; during the trial period it follows Argon pricing at $2 per million input tokens and $10 per million output tokens, adjusted to $4 and $20 after the trial period.

    Same story · 34 reports

    #Model releases#Agents#English text is being prepared.

  4. TechCrunch AIModel releases
    AI score79

    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.

    Same story · 15 reports

    #Model releases#Open-source ecosystem#Safety / alignment

  5. Mistral AIFirst-partyModel releases
    AI score87

    Mistral Releases Open Weights for Mistral Large 4 (Le Chonk)

    Mistral has released the open weights for Mistral Large 4 (Le Chonk), a 1-trillion-parameter natively multimodal model that continues reinforcement learning on top of the preview version, with improved coding and defensive cybersecurity capabilities. It scores 19.7% on FrontierCode 2.0 and 48% on the Artificial Analysis Coding Agent Index, roughly on par with GLM-5.3, but uses an average of 47% fewer tokens per task; on Terminal-Bench 4.0 it improves by 13 percentage points to 41%. The model supports three reasoning modes, uses a modified MIT-based Mistral Ecosystem Open-Weight license, and is priced via API at $0.70 per million input tokens, $0.07 for cached input, and $2.10 for output.

    Le Chonk signal

    Same story · 15 reports

    #Model releases#Open-source ecosystem#Reasoning

  6. cnBetaModel releases
    AI score75

    Google releases Gemini 4 Carbon, tops the Artificial Analysis leaderboard with 59 points

    Google released Gemini 4 Carbon, which scored 59 points on the Artificial Analysis Intelligence Index v5.0 leaderboard, surpassing Claude Opus 5.5's 58 points and GPT-6 Astra's 57 points. The model achieved 88.2%, 54.8%, and 71.0% on SEC-bench Pro, ExploitGym, and CWE-bench, respectively, and Google says it can handle engineering problems in large codebases continuously for several days. The model this week is first being made available to some cybersecurity defenders, then expanded to trusted partners through the Fairwind Program, with no timeline yet for general users; initial pricing is $2 per million input tokens and $10 per million output tokens.

    Same story · 34 reports

    #Model releases#Google#Evaluations / benchmarks

  7. MarkTechPostModel releases
    AI score68

    Google DeepMind Releases Genie 4 Real-Time World Model with Native Audio Generation

    Google DeepMind has released the Genie 4 world model as a research preview to trusted testers; it can generate explorable worlds in real time from text and image prompts, and natively generate ambient sound, music, and character dialogue. The model introduces Gemini 3.8 Flash as a planner within the simulation loop, responsible for breaking down goals, inspecting frames, and replanning, which DeepMind calls an early step toward hybrid world models. Genie 4 is hosted only on Google's platform and will later be opened to AI Ultra users and Gemini API partners; all video and audio carry SynthID watermarks.

    Same story · 28 reports

    #Model releases#Multimodal#Agents

  8. MarkTechPostModel releases
    AI score85

    Google DeepMind releases Gemini 4 Carbon, focused on cybersecurity defense and coding

    Google DeepMind has launched Gemini 4 Carbon, calling it a frontier model for long-horizon agentic coding and cybersecurity defense, ranking first with 59 points on the Artificial Analysis Intelligence Index v5.0. The model achieved 96.2% on the GraphWalks long-context test, and 88.2%, 54.8%, and 71.0% on SEC-bench Pro, ExploitGym, and CWE-bench, respectively, but on OSWorld-2.1 it trailed Claude Opus 5.5's 87.2% with 81.5%. Carbon is currently available only to vetted cybersecurity defenders, with introductory pricing of $2 per million input tokens and $10 per million output tokens, later $4 and $20; the timeline for opening it to developers and the public has not been announced.

    Same story · 34 reports

    #Model releases#Agents#English text is being prepared.

  9. MarkTechPostModel releases
    AI score72

    NVIDIA Releases Nemotron 3.5 Transcribe: An Open-Source 30B MoE Speech Model…

    NVIDIA has released Nemotron 3.5 Transcribe, an open-weight large speech model that supports 40 languages and up to 1 hour of audio per pass, and can select output formats such as verbatim, written, speaker-labeled, contextual, and translation via text instructions. The model has approximately 34B total parameters and about 4.3B activated parameters, adopts a hybrid Mamba-Transformer MoE architecture, can run in BF16 on a single H100 80GB, and the NVFP4 version at about 21GB can run on an RTX 5090. NVIDIA self-reports an average WER of 4.71% across 9 public English Open ASR test sets, the best result among open-weight models, though commercial APIs still lead; its shortcomings are that it only supports offline operation, and the Chinese far-field meeting cpCER is 34.95%.

    Same story · 9 reports

    #Model releases#Speech#Open-source ecosystem

  10. MarkTechPostModel releases
    AI score65

    Block Releases Dessa, a Fully Duplex Speech Model Trained Purely on Speech

    Block's AI research lab has released Dessa, a full-duplex speech model, opening a waitlist in the form of a research preview. It is trained only on human voice conversation audio, without using text or pretrained models. The model can run locally on a MacBook Pro M4 Max or higher configuration, faster than real time; training used one NVIDIA GB200 NVL72 (72 GPUs). No benchmark scores, parameter count, weights, paper, or API have been published so far.

    #Model releases#Speech#Open-source ecosystem

  11. MarkTechPostModel releases
    AI score63

    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.

    Same story · 3 reports

    #Model releases#Open-source ecosystem#Reasoning

  12. Leiphone.comModel releases
    AI score67

    Leiphone.com's In-Depth Breakdown of Google DeepMind's World Model Genie 4

    Google DeepMind has released Genie 4, a new-generation world model that can generate interactive environments in real time and collaborates with Gemini 3.8 Flash through a division of labor: the latter observes the visuals, decomposes tasks, and adjusts plans. Leiphone.com analyzes its technical challenges from three aspects—dynamic language condition injection, Agent causal hallucination, and dual-model asynchronous collaboration—and notes that DeepMind has not yet disclosed the network architecture or scheduling implementation of Genie 4.

    Same story · 28 reports

    #Model releases#Multimodal#Agents

  13. CNETModel releases
    AI score76

    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.

    Same story · 15 reports

    #Model releases#Open-source ecosystem#Agents

Wednesday44 items

  1. TechCrunch AIModel releases
    AI score69

    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.

    Same story · 9 reports

    #Model releases#Speech#Open-source ecosystem

  2. ScienceNetModel releases
    AI score78

    Google DeepMind Releases Research Preview of World Model Genie 4

    On October 27, Google DeepMind released a research preview of its new-generation world model Genie 4, extending the simulation duration for scene consistency to about two minutes—twice that of its predecessor Genie 3—and improving visual realism, physics simulation, and the naturalness of character movements. The model supports real-time changes to weather, viewpoint adjustments, or adding characters and objects via natural language instructions, and features native audio generation capabilities. Google also demonstrated an experiment in which Genie 4, in collaboration with Gemini 4 Argon, simulated Apptronik's Apollo humanoid robot playing checkers, and plans to open it to Google AI Ultra subscribers through Project Genie in the coming weeks.

    Same story · 28 reports

    #Model releases#Multimodal#DeepMind

  3. Analytics India MagazineModel releases
    AI score80

    Google DeepMind Releases Genie 4 Real-Time World Model

    Google DeepMind has released Genie 4, which can generate interactive 3D environments with realistic visuals, sound, and dialogue in real time; it is currently available as a research preview to trusted testers in the creative, news, education, gaming, and engineering fields. The model combines capabilities from the Gemini series, supports modifying scenes, weather, camera angles, and characters via natural language prompts, and natively generates ambient sound and interactive dialogue. Gemini 3.8 Flash acts as a guiding agent in complex simulations, handling task decomposition and progress checks, and all generated content carries a SynthID watermark.

    Same story · 28 reports

    #Model releases#Multimodal#Google

  4. QbitAIModel releases
    AI score56

    Ant Ling Open-Sources Ling-3.1-flash-1019, Intelligence Index Rises to 44

    Ling, the large model under Ant Group, has open-sourced Ling-3.1-flash-1019, which scored 44 on the Artificial Analysis Intelligence Index v4.3.2, a 3-point improvement over the previous version. This version was further trained and optimized on the basis of Ling-3.1-flash, and the base model Ling-3.1-flash-base was also open-sourced, laying the foundation for the knowledge, reasoning, and computational efficiency required for long-horizon tasks.

    Same story · 3 reports

    #Model releases#Open-source ecosystem#Agents

  5. QbitAIModel releases
    AI score60

    SoMark Team Open-Sources 1.3B Document Parsing Model OXR, Topping Three Benchmarks

    The SoMark team has open-sourced the document parsing model OXR. The main model has only 1.3B parameters and topped all three benchmark tests: OmniDocBench, olmOCR-Bench, and DocParsingBench. Its full parsing speed exceeds that of PaddleOCR-VL-1.6 and MinerU2.5-Pro, which were also tested in the same round. The model covers 13 types of document elements, including text, tables, formulas, chemical structure formulas, and code blocks. It can parse reading order and output layout coordinates, converting PDFs and document images into Markdown / JSON for downstream processes such as Agents and RAG.

    Same story · 2 reports

    #English text is being prepared.#English text is being prepared.#Evaluations / benchmarks

  6. QbitAIModel releases
    AI score86

    OpenAI releases GPT-6.1 Terra: API price halved, focused on computer operation capabilities

    OpenAI has released GPT-6.1 Terra, with standard API pricing at $1 per million input tokens, $0.05 per million cached input tokens, and $5 per million output tokens—about half the price of GPT-5.6 Terra. OpenAI says that on Agents' Last Exam, it scores 15 percentage points higher than GPT-5.6 Terra at the medium setting, with per-task cost about one-third of the latter, and matches GPT-6 Astra's performance at one-tenth of its token price; on Terminal-Bench 4.0, it solves 50.3% of tasks at the max setting, compared with 21.5% for the previous generation. Plus, Pro, Enterprise, and Edu users can try it in ChatGPT Work and Codex, and developers can call it via gpt-6.1-terra. On the same day, OpenAI also announced three results in mathematics, computer science, and physics produced by its internal AI system, among which the proof concerning the Exponential Time Hypothesis (ETH) has been formally verified in Lean.

    #Model releases#OpenAI#Agents

  7. QbitAIModel releases
    AI score73

    Baidu Open-Sources ERNIE Unlimited OCR 2, Peak KV Cache Reduced from 47GB to 1.77GB

    Baidu has released and open-sourced ERNIE Unlimited OCR 2. Through the sparse reference sliding window attention mechanism SR-SWA and an infinite decoding mechanism, it reduced the peak KV cache from 47GB under standard decoding to 1.77GB in a single-sequence test at the hundred-page level, a reduction of about 96%. In a 16-way concurrency test, end-to-end throughput increased from 360 to 1315 tokens/second, about 3.65 times that of standard decoding. The model has been open-sourced on Hugging Face and GitHub; the previous generation, Unlimited OCR, has accumulated more than 10 million downloads and about 28,600 GitHub stars.

    Same story · 5 reports

    #Model releases#Open-source ecosystem#Multimodal