New models and versions, open weights, and changes in capability and pricing, with their evaluations.
DeepSeek has launched the experimental multimodal model DeepSeek-V4-Flash-Vision-Exp on its API platform. Its text capabilities match DeepSeek-V4-Flash across agents, reasoning, and world knowledge.
Why it matters: The official release gives the new model's text-capability alignment with V4-Flash and improvements on multimodal-agent benchmarks, helping readers assess its readiness for multimodal agents.
Liquid AI has released DSpark draft-model checkpoints for LFM2.5-1.2B-Instruct, LFM2.5-2.6B, and LFM2.5-8B-A1B. Speculative decoding increases decoding speed without changing output quality, with up to 3.18× higher GPU throughput and up to 2.87× speedup on-device.
Google DeepMind has released Gemini 3.7 Flash, positioning it as its strongest workhorse model for coding and agents, just three weeks after Gemini 3.6 Flash.
Why it matters: The source gives benchmark improvements over Gemini 3.6 Flash and half-price pricing, helping readers weigh cost and performance in coding and agent workloads.
DeepSeek has released DeepSeek-V4-Pro, highlighting upgraded agent capabilities and notable production gains. V4-Pro and V4-Flash offer flexible reasoning levels: low for simple tasks, high for everyday agent workflows, and max for complex tasks. The model natively supports the OpenAI Responses API and is optimized for Codex with one-click configuration. It is now available in the app, on the web in Expert Mode, and through the API; its model name is unchanged.
Why it matters: The official announcement describes V4-Pro's agent upgrades and three reasoning levels, helping readers understand its position in production workflows.
Hugging Face's OlmoEarth Studio can now compute and export embeddings; its code and model weights are public. Through the UI or API, users can select a region, time range, encoder variant (Nano: 128 dimensions and 1.4M parameters; Tiny: 192 dimensions and 6.2M parameters; Base: 768 dimensions and 89M parameters), resolution, and imagery source, then export an int8 Cloud-Optimized GeoTIFF.
Google DeepMind has released SL2T, a multilingual sign-language-to-text model, bringing sign-language AI to consumer products for the first time. On Pixel 11, Gboard and Live Transcribe support dictation from American Sign Language (ASL) to English; more devices and languages are planned.
Why it matters: The source describes the capabilities and limits of a multilingual sign-to-text model and where it is available, helping readers assess sign-language translation in real products.
NVIDIA has released Magpie TTS Multilingual, a 364M-parameter open-weights speech-synthesis model supporting 12 languages. The latest release adds Modern Standard Arabic, Korean, and Brazilian Portuguese, and improves quality in several existing languages through updated training data.
Why it matters: The article reports the 364M open TTS model's language expansion and measured TTFA on different GPUs, helping readers estimate latency budgets for self-hosted voice pipelines.
Meta has released Muse Glimmer, a 30B-parameter multimodal model distilled from Muse and licensed under Apache 2.0. It targets local-agent use cases such as coding, document analysis, and personal assistance.
Why it matters: Meta has open-sourced a 30B multimodal model with reproducible details on its architecture, benchmarks, local deployment, and self-improvement demonstrations.
Google DeepMind has released WeatherNext, an AI model that achieves state-of-the-art forecasts of cyclone tracks, intensity, and wind-field structure, providing about one additional day of lead time on average—roughly a decade of progress in meteorology.
Why it matters: The source reports measured gains of about one additional day of warning for cyclone tracks and intensity and opens the model weights and code.
Mistral AI has released Shieldstral, a 3B open-weight multimodal safety classifier available under Apache 2.0 and runnable on a single 16 GB NVIDIA GPU. It frames content moderation as a policy-adaptive question-answering task: policies are written in natural language at inference time, without retraining. It handles text and images together and outputs calibrated continuous safety scores. Mistral says it matches models up to seven times larger on text safety and sets a new state of the art in multimodal moderation.
Why it matters: The 3B open safety classifier makes moderation policy a natural-language prompt at inference time, offering a view of what smaller models can do for content moderation.
DeepSeek says its official DeepSeek-V4-Flash API is now in public beta. The company reports a major upgrade in agent capabilities, with benchmark scores well above V4-Pro-Preview. It natively supports the Responses API format and is fully adapted for Codex; setup details are in the official API documentation.
Why it matters: DeepSeek has announced the V4-Flash API beta and improved agent capabilities, offering a basis for assessing integration with Codex.
Mira Murati says Inkling-Small is comparable to Inkling at a quarter the size, with open weights and fine-tuning on Tinker today. She looks forward to seeing what people make with it. In the quoted announcement, @thinkymachines reports 276B total parameters and 12B active, makes the full weights available, and offers text, image, and audio chat on Tinker Playground. https://thinkingmachines.ai/news/inkling-small/
Why it matters: The comparison with Inkling provides context for the smaller model’s claimed performance, while the parameter counts clarify the scale of the size reduction.
Google DeepMind has released Gemini Robotics ER 2, positioned as the robot's high-level brain: it converses with people, understands the physical world, and plans multi-step tasks before handing execution to a lower-level VLA model. It can also call tools such as Google Search natively.
Why it matters: The official release gives concrete figures for ER 2's progress classification and temporal localization, along with access details, helping readers assess its limits on long tasks.
Google DeepMind has released Lyria 3.5, its next-generation music-generation model, in Google Flow Music. The release improves musicianship, lyrics, vocals, and creative control: it produces richer, more natural melodic structures; improves lyric quality, prompt adherence, and structure awareness; offers more expressive vocals with finer emotional detail and better pronunciation; and makes tempo and duration easier to control.
Google DeepMind has released Gemini Robotics 2 with three models: Gemini Robotics 2, a vision-language-action model for translating visual and language input into motion; Gemini Robotics ER 2 for embodied reasoning; and Gemini Robotics On-Device 2, which runs locally on devices.
Why it matters: The article explains the three models' roles and access points, helping readers assess pathways to full-body robot control and multi-robot collaboration.
NVIDIA has released Cosmos-H-Dreams, a real-time, action-conditioned generative simulator for surgical robots. It distills Cosmos-H-Surgical-Simulator into a causal few-step student model and runs through the FlashDreams inference library on a single NVIDIA RTX PRO 6000 GPU.
Google DeepMind has released three models: Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber for security work.
Why it matters: The source compares token efficiency, pricing, and benchmark results across the three Flash models, helping readers assess cost changes for agent workflows.
Google DeepMind has released Gemini 3.5 Flash Cyber, a lightweight cybersecurity model fine-tuned from 3.5 Flash to quickly find, validate, and patch vulnerabilities. Across benchmarks such as CyberGym, it approaches the results of larger models with multiple calls.
Why it matters: The source reports the lightweight security model's performance on benchmarks such as CyberGym and the terms of its limited pilot, helping readers assess its positioning and availability.
Inkling is Lilian Weng’s organization’s open weights model. It aims to serve as a foundation with solid performance across broad categories of capabilities, for practical use and customization. Weng invites users to try it on Tinker.
Why it matters: The release positions Inkling as a foundation for practical customization, with full weights and fine-tuning availability giving concrete context for its intended use.
Thinking Machines Lab has released the open-source Inkling family on Hugging Face. Inkling is an MoE model with about 1T total and 41B active parameters, natively supporting image, audio, and text inputs with a 1M-token context. The release also includes Inkling-Small, with 276B total and 12B active parameters.
Why it matters: The article gives Inkling's architecture, deployment setup, and benchmark comparisons, helping readers assess the practical requirements of an approximately 1T-parameter open multimodal model.
Mistral AI has released Robostral Navigate, its first embodied-navigation model. The 8B model uses a single standard RGB camera, with no LiDAR or depth sensor, and achieves a 76.6% success rate on R2R-CE validation unseen—9.7 percentage points above the best single-camera approach and 4.5 points above the best systems using depth or multiple cameras.
Why it matters: The source reports navigation success using a monocular RGB camera and describes the training approach, giving a view of how sensor requirements for embodied navigation are changing.
Mistral has released Leanstral 1.5, a formal-verification model licensed under Apache 2.0 with 119B total and 6B active parameters. It scores 100% on miniF2F, solves 587 of 672 PutnamBench problems, and sets state-of-the-art scores of 87% on FATE-H and 34% on FATE-X.
Why it matters: The source reports Leanstral 1.5's formal-verification benchmark results and open-source availability, helping readers assess its usefulness in proof engineering.
Google DeepMind has released Nano Banana 2 Lite and Gemini Omni Flash. The former is the fastest model in the Nano Banana family.
Why it matters: Both models are available to developers, with details on latency, pricing, and capability limits to help assess the cost and usability of the image-to-video pipeline.
Google has released TabFM, a zero-shot foundation model for classification and regression on tabular data. It recasts table prediction as an in-context-learning problem, producing predictions in a single forward pass without manual training, tuning, or feature engineering.
Why it matters: The source details TabFM's architecture and TabArena results, helping readers assess whether zero-shot table prediction can replace traditional tree-model workflows.
Mistral has released OCR 4, which returns bounding boxes, block-type classifications, and per-page, per-word confidence scores in addition to extracted text. It supports 170 languages across 10 language families and can be fully self-hosted in a single container.
Why it matters: OCR 4 returns bounding boxes, block classification, and confidence scores, helping readers assess how document parsing can fit into RAG and agent workflows.
Google DeepMind has released DiffusionGemma, an experimental open-source model that uses text diffusion to generate entire chunks of text in parallel. It generates text up to 4× faster on specialized GPUs, exceeding 1,000 tokens per second on a single NVIDIA H100 and 700 tokens per second on an RTX 5090.
Why it matters: Google DeepMind's open text-diffusion model comes with concrete speed and hardware figures, helping readers assess the feasibility of local interactive inference.
Andrej Karpathy says Claude Fable 5 shares the same underlying model as Mythos but adds safety safeguards, and reaches state-of-the-art results by some margin on almost every benchmark.
Why it matters: Drawing on firsthand use, Karpathy calls Fable 5 a release worthy of a major version-number jump and describes concrete changes on long-horizon tasks.
Google DeepMind has released Gemini 3.5 Live Translate, a near-real-time speech-to-speech translation audio model supporting more than 70 languages. It detects languages automatically and preserves the speaker's intonation, rhythm, and pitch.
Why it matters: Gemini 3.5 Live Translate replaces sentence-by-sentence waiting with continuous generation and offers developer, enterprise, and consumer deployment paths.
Google DeepMind has released Gemma 4 12B, an agentic multimodal model aimed at laptops, positioned between the edge-focused E4B and the 26B MoE. It is the company's first medium-sized model with native audio input.
Why it matters: Gemma 4 12B removes multimodal encoders and feeds visual and audio input directly into the LLM, giving readers a view of hardware requirements for local multimodal agents.
Google Research has open-sourced its hydrology framework on GitHub under Apache 2.0, enabling national weather and hydrology agencies to train AI flood-forecasting models using the same architecture and similar training data as Google Flood Hub.
Mistral has released Mistral Medium 3.5, a dense 128B model with open weights under a modified MIT license. It becomes the default model in Mistral Vibe and Le Chat, with API pricing of $1.50 per million input tokens and $7.50 per million output tokens.
Why it matters: Mistral is bringing its coding agent to the cloud while opening the weights of a 128B model, giving readers a basis for judging the feasibility of self-hosted coding agents.
At Google I/O, Oriol Vinyals announced the launch of Gemini 3.5 Flash, describing the goal as frontier-level intelligence at unprecedented speed. He says 3.5 Flash beats 3.1 Pro on almost every benchmark while running at Flash speeds.
Why it matters: The post states the new model's positioning and benchmark claim against 3.1 Pro, giving a quick read on Google's Flash-tier speed and capability tradeoff.
Google DeepMind has released Gemini Omni Flash, the first model in the Omni family. It can combine images, audio, video, and text as inputs to generate high-quality video and supports iterative video editing through natural-language conversation.
Why it matters: The source describes the first Omni model's input and output capabilities and where it is available, helping readers assess the practical reach of multimodal video generation.
Google DeepMind has released Gemini 3.5 Flash, the first model in the Gemini 3.5 family, positioning it as a frontier model for agents and coding. It is available starting today in the Gemini app, AI Mode in Google Search, Google Antigravity, the Gemini API, and Gemini Enterprise.
Why it matters: Gemini 3.5 Flash beats 3.1 Pro on coding and agent benchmarks and includes speed and cost comparisons, helping readers assess its potential for long-horizon tasks.
Nick Turley welcomes the Instant update, saying hundreds of millions of people use this version of ChatGPT by default and highlighting improvements in intelligence, concision, accuracy, and conversation quality. In the quoted post, @michpokrass says gpt-5.5 instant shipped today and is rolling out to everyone over the next couple of days, describing gains in factuality, “crushing hacks,” and baseline intelligence, with significantly fewer hallucinations.
Why it matters: Instant’s role as the default ChatGPT version for hundreds of millions of people gives the reported improvements in accuracy and conversational quality a broad everyday audience.
Google DeepMind has released Gemini 3.1 Flash TTS, a text-to-speech model focused on control, expressiveness, and audio quality. It is available in developer preview through the Gemini API and Google AI Studio, in enterprise preview on Vertex AI, and to Workspace users through Google Vids.
Why it matters: The official release explains how to use audio tags and lists language coverage, helping readers assess how controllable speech generation is.
Google DeepMind has released Gemini Robotics-ER 1.6, a major upgrade to its reasoning-first robotics model. It improves spatial reasoning and multi-view understanding and adds the ability to read complex instruments such as pressure gauges and level meters.
Why it matters: The source details changes in spatial reasoning, multi-view understanding, and instrument reading, along with access information, giving a view of progress in high-level robot reasoning.
Google DeepMind has released the open Gemma 4 family in four sizes: E2B, E4B, 26B MoE, and 31B Dense, under the Apache 2.0 license.
Why it matters: Gemma 4 spans edge devices to workstations in four sizes and reports Arena rankings under Apache 2.0, helping readers assess the practical range of open models.
Mistral AI has released Voxtral TTS, its first text-to-speech model. The 4B model supports nine languages—English, French, German, Spanish, Dutch, Portuguese, Italian, Hindi, and Arabic—and offers expressive emotion and low latency.
Why it matters: The source reports the 4B model's language coverage and 70 ms latency, and compares it with ElevenLabs, helping readers assess cost and usability for voice agents.
Mistral AI has released Mistral Small 4, the next major version of the Mistral Small family and its first model to combine Magistral's reasoning, Pixtral's multimodal, and Devstral's agentic-coding capabilities in one model. It is licensed under Apache 2.0.
Why it matters: Mistral has combined reasoning, multimodal, and coding capabilities in one open small model, with adjustable reasoning effort and deployment requirements to help assess implementation costs.
