Search “NVIDIA”

Google DeepMind has released Gemini 4 Carbon, which ranks first on the Artificial Analysis Intelligence Index v5 with a score of 59 and leads in 4 of 9 evaluations including HLE-Diamond and TB-Science. The model focuses on cybersecurity and long-horizon tasks, claiming 88.2% on SEC-bench Pro, 54.8% on ExploitGym, and 71.0% on CWE-bench, and is initially available only to trusted cybersecurity personnel through the Fairwind Program. In terms of specifications, it supports a 1M token context and 1

Matched fieldsBody

The Baseten engineering team, through native audio decoding, KV cache FP8 quantization, and reduced host-device synchronization, increased batch transcription throughput for Qwen3-ASR 1.7B on a single RTX PRO 6000 by 7x and reduced latency by 6.9x. After optimization, the transcription cost for one hour of audio is approximately $0.0013, i.e., 1/10 of a cent per audio hour, with no significant change in word error rate. The optimized version has been made the default deployment in the Baseten Mo

Matched fieldsBody

In his long-form article, Ed Zitron analyzes Anthropic's S-1, arguing that the document only discloses Q2 2026 data, and that the quarter's $974 million in profit came from two months of SpaceX compute discounts, while Q3 revenue, cash balance, and payment schedule were not disclosed. He tallies Anthropic's non-cancelable compute commitments with Google, Amazon, Microsoft, SpaceX, Broadcom, and others at approximately $269.4 billion, with an additional $125 billion to $150 billion in TPU purchas

Matched fieldsBody

Iluvatar CoreX held its first live demonstration of the Tianyao 300 at a media open day on October 28, claiming that in comparative tests on large language models such as GLM-5.3 as well as text-to-image and text-to-video scenarios, it outperformed the H200 overall. On the training side, in an 8-GPU full-system FP8-precision demonstration based on Qwen3-32B, the Tianyao 300 achieved per-GPU effective throughput 1.33 times that of the H100. The Tianyao 300 was unveiled at this year's World Artifi

Matched fieldsBody

Google has released Gemini 4 Carbon, a frontier model aimed at cybersecurity and long-horizon coding tasks, topping 15 of Google's own 17 benchmarks, but it has not yet been made public and is available only to some cybersecurity defenders. It scores 59 on Artificial Analysis' Intelligence Index, surpassing Claude Opus 5.5 and up 9 points from Argon four weeks ago; on long-input reasoning it leads OpenAI Astra and Opus by 24 points. The introductory API price is $2 per million input tokens and $

Matched fieldsBody

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 sup

Matched fieldsBody

Microsoft's capital expenditure (including finance leases) for the September quarter was $45.2 billion, $6.5 billion less than analysts' expected $51.7 billion, but still up about 30% year-over-year. CFO Amy Hood attributed the shortfall to delays in third-party data center capacity deliveries, with the related facilities expected to come online in the second half of Microsoft's fiscal year, while also lowering its calendar 2026 spending forecast from $175 billion to $170 billion. Microsoft adde

Matched fieldsBody

On October 28, MiniMax open-sourced MCode Stack under the MIT license, including the 17 SKILL.md skills and 5 checklists its team used to build MiniMax Code, but not the MiniMax Code source code. The skills are grouped into planning and decomposition, context and tools, long sessions and commands, review, and experiments and hotfixes, among which independent-review provides reviewers only with the diff, requirements, and test results. It can be installed into MiniMax Code by copying the skill di

Matched fieldsBody

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, w

Matched fieldsBody

AssemblyAI has released its first text-to-speech model, AssemblyAI-TTS-1, and officially opened it up, targeting real-time telephony voice agents, priced at $15 per million characters, with no annual contracts and no concurrency tiers. The model is offered only as a hosted API in us-east-1 and eu-north-1, supports HTTP, WebSocket and OpenAI-compatible interfaces, and comes with more than 4,000 voices, instant voice cloning and 6 languages. The vendor's own testing shows 93.7% entity accuracy and

Matched fieldsBody

Infini-AI-Lab has released the open-source framework Hermit, which adapts AI agents by training external harness adapters while keeping the model and harness frozen; v0.1.0 was released on October 25, 2026. The framework is implemented in Python under the Apache-2.0 license and can be installed via pip install hermit-harness, and it includes five evolution methods: GEPA, SkillOpt, MetaWorld Model, Harness², and Continual Harness. On GLM 5.3 Flash, the MWM adapter achieved 0.795 accuracy on the C

Matched fieldsBody

Google DeepMind has released Genie 4, a new-generation world model that can turn prompts into explorable, realistic scenes in real time, and for the first time the Genie model itself generates background sounds and music, while characters in the scene can also hold real-time conversations. Users move via the keyboard and use text prompts to modify content in the scene instantly, with the scene remaining coherent across multiple inputs. In the demo, a simulated Apptronik Apollo robot plays checke

Matched fieldsBody

LimX Dynamics has updated the second version of the technical report for its open-source embodied engineering foundation FluxVLA Engine. Compared with the first version in September, it newly adds post-training experiments with online reinforcement learning and expands the coverage of simulation evaluations and real-robot deployment. The report shows that after online reinforcement learning, SmolVLA's success rates on the ten tasks of LIBERO-10 and RoboTwin Clean increased by 10.0 and 15.4 perce

Matched fieldsBody

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

Matched fieldsBody

Matched fieldsBody

Google's TPUs will run on a hyperscale cloud outside Google for the first time, with Oracle deploying capacity starting in calendar year 2027 and gradually expanding over several years. Oracle provides power, cooling, and data center networking; its Acceleron network connects TPU clusters to Oracle's databases and storage, letting enterprises fine-tune and deploy models next to their data without cross-cloud transfers. Google says the eighth-generation TPU delivers up to 12.6 PFLOPS of FP4 compu

Matched fieldsBody

OpenBMB has released the open-source MiniCPM-OCR suite, comprising the 1B MiniCPM-OCR 1.0 and the 2B decision model MiniCPM-OCR Jev. MiniCPM-OCR 1.0 scores 93.68 on Real5-OmniDocBench and 54.7 on OCRBench v2; Jev can perform cross-model verification of text, formulas, and tables, scoring 93.6 on JevOCRBench, surpassing JEV-35B-A3B's 90.1 and approaching Qwen3.5-397B-A17B's 93.8. Powered by DSpark, its speed is 1.49 times that of PaddleOCR-VL-1.6 and 3.62 times that of MinerU2.5 Pro; with SDK con

Matched fieldsBody

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 Liqu

Matched fieldsBody

Prime Intellect has released Prime Mega MoE, which builds on the Mega MoE kernel from DeepSeek's open-source DeepGEMM library by adding a fused and overlapped backward pass, making it usable for training rather than inference only. In the prime-rl trainer, the full training step for Qwen3 MoE models is 32% to 34% faster than the baseline and 48% to 55% faster than TorchTitan; the single-layer forward pass is up to 3.11x faster (from 12.0 ms to 3.9 ms), but the backward pass for Qwen3-235B-A22B i

Matched fieldsBody

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 200

Matched fieldsBody