60 releases covered live on the show — every model, product, paper and tool that mattered, with links and our analysis.
Qwen launches Omni 7B: sees, hears, reads, and talks back
Qwen released Qwen2.5-Omni-7B, an open-weights omni-modal model that perceives text, images, audio, and video, and generates both text and speech. It packs end-to-end multimodal perception and spoken output into a 7B parameter model available on Hugging Face.
DeepSeek silently drops V3-0324, 685B params under MIT license
DeepSeek silently updated their V3 base model with DeepSeek-V3-0324, a 685B parameter MoE released on Hugging Face under the MIT license. This is not R1 (their reasoning model) but the powerful base model R1 was built on, and supposedly the base for a future R2.
Google reclaims #1 with Gemini 2.5 Pro thinking model
Google dropped Gemini 2.5 Pro, a thinking model that took the #1 spot as the best all-around LLM available, with massive jumps on benchmarks like AIME (up nearly 20 points) and GPQA. It inherits native multimodality and a 1M token context window, maintaining high accuracy even at 120k+ tokens on needle-in-a-haystack tests, with surprisingly low latency (~13 seconds on hard reasoning questions vs 45+ for others). Tulsee Doshi, head of product for Gemini models, joined the show to give the inside scoop.
Ideogram 3.0 launches with strong text, logos, and style references
Ideogram launched version 3.0 of its image generation model with another SOTA claim. It is particularly strong on text and logo rendering, photorealism, and style references, continuing Ideogram's edge in typography-heavy image generation.
GPT-4o gets an update, ties for #1 on LMArena beating GPT-4.5
OpenAI shipped a new GPT-4o checkpoint (2025-03-26) that jumped over GPT-4.5 to tie for #1 on LMArena. The update landed as the show was being written, read as a direct response to Gemini 2.5's launch in the escalating frontier-model race.
Reve emerges with SOTA diffusion image generation claims
Reve launched a new diffusion image generation model claiming state-of-the-art quality, reportedly beating heavyweights like Midjourney and Flux at roughly a penny per image. The previously low-profile lab made a splash with strong prompt adherence and image quality.
Canopy Labs drops Orpheus 3B natural-sounding speech model
Canopy Labs released Orpheus, an open speech language model that produces natural, human-sounding speech, headlined by a 3B model with smaller variants (1B, 500M, 150M) in the family. Weights are on Hugging Face with a Colab for trying it out, discussed on the show with Daily.co CEO Kwindla Kramer in the voice AI segment.
LG open sources EXAONE and EXAONE Deep 32B reasoning model
LG AI Research open sourced its EXAONE family, headlined by EXAONE Deep 32B, a thinking/reasoning model. The release puts a large Korean lab's reasoning model in open weights on Hugging Face, and Alex published a live reaction video to the launch.
Mistral Small 3.1 24B: open-weights multimodal model
Mistral released Mistral Small 3.1, a 24B-parameter open-weights model that adds multimodal (vision) capabilities to the Small line. Both instruct and base checkpoints were published on Hugging Face, making it a strong local multimodal option at the 24B size class.
NVIDIA Canary Flash: Apache 2 speech recognition and translation
NVIDIA released Canary 1B Flash and 180M Flash, Apache 2.0 licensed speech recognition and translation models built as Llama finetunes. The permissive license makes them freely usable for commercial ASR and translation workloads.
NVIDIA drops Llama-Nemotron reasoning models plus training dataset
NVIDIA released the Llama-Nemotron family, including Super 49B and Nano 8B reasoning models, announced around GTC. Alongside the open weights, NVIDIA published the Llama-Nemotron post-training dataset, giving the community both the models and the data recipe behind them.
OpenAI launches steerable voice model and two new transcription models
OpenAI launched a new emotionally steerable text-to-speech voice model plus two new transcription models, watched live on the show as a watch party. The TTS model can be instructed how to speak (tone, emotion, character), demoed at openai.fm, and the models are available through the API for voice agents.
Roboflow drops RF-DETR, a SOTA open-source object detection model
Roboflow released RF-DETR, a state-of-the-art real-time object detection model, announced as breaking news on the show by CEO Joseph Nelson. The model is fully open source on GitHub and targets practical, deployable computer vision workloads.
StepFun releases Step-Video-TI2V image-to-video model
Chinese lab StepFun dropped Step-Video-TI2V, an open text/image-to-video generation model. Weights are on Hugging Face with code on GitHub, adding another open-weights option to the fast-moving video generation space.
Tencent updates Hunyuan3D 2.0 with MultiView and Turbo variants
Tencent updated its Hunyuan3D 2.0 image-to-3D model with an MV (MultiView) version that conditions on multiple input views, plus a faster Turbo variant. The show highlighted it as new SOTA for 3D generation, available to try in a Hugging Face space.
AllenAI ships OLMo 2 32B, a fully open GPT-4-class model
The Allen Institute for AI released OLMo 2 32B, its biggest fully open model yet, with weights, code, and dataset all published under Apache 2.0. Announced by Nathan Lambert as a last-second addition, it reportedly beats GPT-3.5 and GPT-4o mini as well as leading open-weight models like Qwen and Mistral at its size.
ByteDance unveils Seedream 2.0 bilingual image generation foundation model
ByteDance released Seedream 2.0, a native Chinese-English bilingual image generation foundation model, alongside a technical paper. It emphasizes excellent text rendering (especially Chinese), cultural nuance, and human preference alignment, generating high-quality, culturally relevant images from prompts in either language.
Cohere Command A: 111B enterprise model with 256K context on just 2 GPUs
Cohere announced Command A, a 111B parameter open-weights model with a 256K context window, presented on the show by Cohere's Sandra Kublik. It runs on only two GPUs where models of this size typically require around 32, and is built for enterprise use: agentic tasks, tool use, multilingual performance, and secure private deployments.
EuroBERT: multilingual encoder models from 210M to 2.1B parameters
EuroBERT is a new family of multilingual encoder models ranging from 210M to 2.1B parameters, trained on a 5 trillion-token dataset across 15 languages with 8K context support. It targets European and global language NLP tasks like retrieval and RAG, where properly encoding non-English character sets matters.
Google open sources Gemma 3, 1B-27B multimodal family with 128K context
Google released Gemma 3, an open-weights model family spanning 1B to 27B parameters with multimodal (text, image, video) capabilities, support for over 140 languages, and a 128K context window. The 27B model runs on a single GPU, with Sundar Pichai claiming competitors need roughly 10x the compute for similar performance. It shipped with day-one open source ecosystem support (Hugging Face, Ollama, Kaggle) plus ShieldGemma 2 for content moderation.
OpenSora 2.0: 11B open-source video model trained for $200K
OpenSora 2.0 is an 11B parameter open-source video generation model that claims state-of-the-art results while costing only about $200,000 to train. The team claims performance approaching OpenAI's Sora on some benchmarks, underscoring how fast open-source video generation is improving.
Nous Research releases DeepHermes 24B and 3B hybrid reasoning models
Nous Research released DeepHermes hybrid reasoners at 24B (Mistral-based) and 3B sizes, models that can toggle between standard chat responses and long chain-of-thought reasoning. The 24B preview is available on Hugging Face as part of the week's wave of open-source reasoning model releases.
Reka Flash 3: 21B open-source reasoning model under Apache 2.0
Reka AI open sourced Reka Flash 3, a 21B parameter reasoning model released under an Apache 2.0 license and trained with the REINFORCE Leave One-Out (RLOO) reinforcement learning technique. It excels at chat, coding, instruction following, and function calling, with Nisten calling it possibly one of the best ~20B models available.
Remade AI releases 8 open LoRA video effects for Wan 2.1
Remade AI published eight LoRA video effects for Alibaba's Wan 2.1 14B image-to-video model, including effects like squish, inflate, deflate, and cakeify. The open release shows video effects becoming trainable and customizable via LoRAs on top of open video models.
AI21 releases Jamba 1.6 Large and Jamba 1.6 Mini open-weights models
AI21 Labs released Jamba 1.6 in Large and Mini sizes, updating its hybrid SSM-Transformer (Mamba-based) model family with open weights on Hugging Face. The Jamba architecture targets long-context efficiency compared to pure transformer models.
Qwen releases QwQ-32B reasoning model that matches R1 on some evals
Alibaba's Qwen team released QwQ-32B, an open-weights reasoning model that matches DeepSeek R1 on several evals despite being roughly 20x smaller at 32B parameters. Qwen tech lead Junyang Lin joined the show to announce it, and the episode dubbed it Alibaba's 'R1 killer' for bringing strong reasoning to a size that runs on consumer hardware.
Cohere For AI releases Aya Vision 8B and 32B open multilingual vision models
Cohere For AI released Aya Vision in 8B and 32B sizes, extending the multilingual Aya family with open-weights vision-language capabilities. The models target multilingual multimodal understanding across many languages.
NotaGen open symbolic music model generates classical sheet music
NotaGen is an open symbolic music generation model that produces high-quality classical sheet music rather than raw audio. The release includes code on GitHub, weights on Hugging Face, and a browser demo.
MiniMax launches Image-01 text-to-image model at 1/10 the cost
MiniMax released Image-01, a versatile text-to-image model the company positions at roughly one tenth the cost of competing image generation offerings. It is available through MiniMax's hosted platform.
Tencent releases HunyuanVideo-I2V open image-to-video model
Tencent finally shipped the long-awaited image-to-video version of HunyuanVideo, with open weights on Hugging Face and a hosted try-it experience. It lets users animate still images using one of the strongest open video generation models.
Zhipu AI open-sources CogView 4, a 6B text-to-image model
Zhipu AI released CogView 4, a 6B-parameter open text-to-image model in the CogView family, with code available on GitHub. It is notable as an open-weights image generation option with strong Chinese and English prompt support.
Arcee AI announces Conductor, an intelligent model router
Arcee AI's Lucas Atkins joined the show to announce Conductor, a model router that picks the best model (including Arcee's small specialized models) for each query. It targets cost and quality optimization by routing requests instead of sending everything to one large model.
Manus AI research agent has everyone talking
Manus is a new AI research agent (manus.im) that creates a to-do list, browses the web in a real Chrome browser, and generates files, described on the show as 'Operator on steroids' and seemingly powered by Claude 3.7 behind the scenes. The crew tested it live on a research task and praised its slick UI.
Elysian Labs launches Auren iOS app
Elysian Labs (from nearcyan) launched Auren, an iOS app offering an emotionally attuned AI companion experience. The launch drew attention for its polished consumer approach to AI companionship.
Google announces AI Mode in Search powered by Gemini 2.0
Google announced AI Mode, a new conversational search experience in Google Search, alongside Gemini 2.0-powered upgrades to AI Overviews. Robby Stein, VP of Product for Google Search, joined the show for an exclusive interview about the launch, which brings full AI chat-style answers with follow-ups directly into Search.
Sesame's ultra-realistic conversational voice demo takes the world by storm
Sesame released a demo of its conversational speech model featuring the Maya voice, and its naturalness, with human-like pauses, laughs, and interruptions, went viral across the AI community. Alex recorded a reaction conversation with Maya showcasing how lifelike the voice model is.
OpenAI updates ChatGPT advanced voice mode with semantic VAD
Alongside the image generation launch, OpenAI quietly updated ChatGPT's advanced voice mode with semantic voice activity detection. The model now understands when you have actually finished speaking rather than cutting in on pauses, leading to much more natural conversation flow.
OpenAI enables native image generation in GPT-4o, internet goes Ghibli
OpenAI finally enabled GPT-4o's native auto-regressive image generation in ChatGPT, sparking the biggest mainstream AI buzz of the week as the internet ghiblified itself. Launched right after Gemini 2.5, it excels at instruction following, text rendering, and multi-turn editing, with viral demos ranging from ad mockups to a full Lord of the Rings trailer.
OpenAI adopts Anthropic's Model Context Protocol - MCP won
OpenAI officially announced support for the Model Context Protocol (MCP) in its Agents SDK, effectively settling the agent tool-connectivity standards war in MCP's favor. Possibly more impactful long-term than the week's flashier launches, since the entire ecosystem can now converge on one protocol for connecting models to tools and data.
Cursor ships Claude 3.7 MAX mode
Cursor shipped Claude 3.7 MAX, a mode giving the agent the full context window and higher tool-call limits with Claude 3.7 Sonnet. It is aimed at harder, longer coding tasks at premium usage-based pricing.
Google makes Deep Research free, adds Canvas and Live Previews to Gemini
Google made its Deep Research agent free for Gemini users and shipped Canvas, a collaborative workspace with live previews for code and documents. Demos on the show included a playable Tetris game and a markdown word counter built and previewed directly inside Gemini.
NotebookLM teases Mind Maps for visualizing sources
Google's NotebookLM team previewed Mind Maps, a feature that turns your uploaded sources into interactive visual maps of concepts. It was teased publicly by the team this week ahead of a wider rollout.
Google AI Studio adds native YouTube video understanding via link dropping
Google AI Studio now lets you drop a YouTube link and have Gemini natively understand the video. This unlocks video analysis, summarization, and support use cases without downloading or preprocessing the content.
Google makes Deep Research free in the Gemini app, powered by Gemini Thinking
Google made its Deep Research agent free for everyone in the Gemini app and upgraded it to run on Gemini Thinking. In a live test on the show it browsed over 150 websites to compile a comprehensive answer, with a polished interface and export to Google Docs.
Gemini Flash gains native image generation and conversational editing
Google enabled native image generation in Gemini Flash Experimental, letting users generate and iteratively edit images conversationally inside the same multimodal model. The crew demoed it live on stream, editing photos of themselves with natural-language instructions, and saw it as a preview of how creative tools like Photoshop will work.
Grok Voice mode opens up to free users
xAI made Grok's voice mode available to free users, removing the paid-tier requirement. The expansion brings conversational voice AI to everyone on the Grok app.
OpenAI makes o1-pro available via API at $600 per 1M output tokens
OpenAI exposed its o1-pro reasoning model through the API for the first time, priced at $600 per million output tokens. The show jokingly framed the pricing as 'for oligarchs', but it makes OpenAI's highest-compute reasoning tier programmatically accessible.
Nous Research opens Portal, an inference API for Hermes models
Nous Research launched Portal, its new inference API service offering access to models like Hermes 3 Llama 70B and DeepHermes 3 8B directly via API. It marks another open-source lab standing up hosted API access to make its models more accessible.
OpenAI launches Responses API with Web Search, File Search, and Computer Use
OpenAI announced a new agent-focused developer stack at a livestream: the Responses API, a new way to build with OpenAI designed for agentic workloads, plus an Agents SDK. It ships with three built-in tools: Web Search, a File Search tool providing built-in RAG over your files, and a Computer Use tool for agents that operate computer interfaces.
Mistral announces state-of-the-art OCR API
Mistral AI announced Mistral OCR, a document-understanding API the company claims is state of the art at extracting text, tables, and equations from complex documents. It targets RAG and document-processing pipelines with structured markdown output.
Prince Canuma releases MLX-Audio v0.0.3 for speech on Apple Silicon
Prince Canuma, creator of MLX-VLM, FastMLX, and MLX Embeddings, released MLX-Audio v0.0.3, an open-source library bringing speech and audio models to Apple Silicon via MLX. It makes powerful open-source TTS and audio models accessible locally on Mac hardware.
W&B ships official Weave MCP server - talk to your evals
Weights & Biases shipped an official MCP server for Weave, its LLM observability and evaluation tool, letting agents and MCP clients query and analyze your evals directly. Morgan McQuire of the W&B Applied AI team demoed it on the show, with wandb Models integration coming soon so agents can monitor loss curves for you.
Gemini Co-Drawing demo uses native image output to help you draw
A Hugging Face space demo, Gemini Co-Drawing, uses Gemini's native image generation output to collaboratively complete and enhance your sketches as you draw. It showcases the new native image-output capability of Gemini 2.0 Flash in an interactive tool.
Baidu launches Miaoda no-code AI app building tool
Baidu introduced Miaoda, a no-code AI-powered build tool that lets users create applications without writing code. It joins the growing wave of AI-assisted app builders coming out of Chinese tech giants.
Cloudflare ships support for building MCP servers on Workers
Cloudflare published tooling and docs for building and deploying Model Context Protocol servers on Cloudflare Workers, riding the MCP wave sweeping the AI community. Senior PM Dina Kozlov joined the show's MCP deep dive to walk through it alongside MCP builder Jason Kneen.
Google ships Gemini-powered Data Science Agent in Colab
Google launched a Data Science Agent inside Google Colab, powered by Gemini, that can autonomously generate complete, working notebooks from natural language descriptions of an analysis task. It automates data loading, exploration, and modeling boilerplate for data scientists.
ByteDance releases DAPO, an RL method that beats GRPO
ByteDance published DAPO, a reinforcement learning method for LLM post-training presented as an improvement over GRPO. The paper ships with an open GitHub implementation, making the technique reproducible for the open-source RL community.
ARC-AGI 2 benchmark revealed, thinking models score just 4%
The ARC Prize Foundation revealed ARC-AGI 2, the next iteration of the abstract reasoning benchmark. Base LLMs score 0% and even thinking models only reach about 4%, showing how far current frontier models remain from human-level fluid intelligence.
Roboflow launches RF100-VL benchmark for vision-language models
Alongside RF-DETR, Roboflow introduced RF100-VL, a new evaluation benchmark for vision-language models built from real-world detection datasets. It gives the community a grounded way to measure how well VLMs handle practical object detection tasks.
Weights & Biases is acquired by CoreWeave
CoreWeave announced it is acquiring Weights & Biases, the AI developer platform and ThursdAI's home company. The deal pairs W&B's experiment tracking, Weave, and models tooling with CoreWeave's AI cloud infrastructure.
Don't wait a month — get every week's releases recapped in your inbox each Thursday.
