分类筛选
找到 1605 个 Demo
New AI models drop almost every day of the week....
New AI models drop almost every day of the week. But let’s compare Gemini 3.1 with other same-tier models—different DNA. → Gemini 3.1 Flash-Lite 389 t/s, $0.25/M input. Built for high-volume, at-scale workloads. 86.9% GPQA Diamond. 1M context window. → Claude Haiku 4.5 101 t/s, $1/M input. 73.3% SWE-bench. Near-frontier coding at a budget price. Best for Anthropic-ecosystem sub-agents. → GPT-5.3 Instant ~$0.80/M input. 26.8% fewer hallucinations with web search vs. the prior model. Faster, more direct answers. See the full breakdown in the infographics ↓
Here we go: Gamma 4 released: ""Outperforms models...
Here we go: Gamma 4 released: ""Outperforms models 20x its size" Google dropped Gemma 4 under Apache 2.0, full open-source, big licensing shift. Built on Gemini 3 tech, four sizes: E2B, E4B, 26B MoE, 31B Dense. Price-performance: 31B is #3 open model on Arena AI, 26B MoE is #6 — beating models 20x their size. MoE activates only 3.8B params at inference. Fits on consumer GPUs quantized. Edge: E2B/E4B run offline on phones, Raspberry Pi, Jetson Nano. Native vision + audio at 2B params, 128K context. Built with Qualcomm/MediaTek.
Introducing SenseNova-SI-1.4-InternVL3-8B: scaling...
Introducing SenseNova-SI-1.4-InternVL3-8B: scaling spatial intelligence across open multimodal models. 🚀 ⚡ 8M curated samples: SenseNova-SI-8M — rigorous taxonomy of spatial capabilities 🧠 SOTA spatial reasoning: hits 88.8 on MindCube-Tiny, 40.1 on MMSI — beats Gemini-2.5-Pro & GPT-5 🎯 Grounding: 89.21 avg on RefCOCO splits, 78.64 on CountBench 📐 Depth estimation: 95.56 relative depth, 80.31 absolute depth on Ibims 🛠️ Multi-base: InternVL3, Qwen3-VL, Qwen2.5-VL, BAGEL (unified understanding + generation) ✅ Apache 2.0. Drop-in for existing research pipelines. 🤖 Model: https://t.co/bxG2j4a3Qq 📄 Paper: https://t.co/HCtbzPfS5r
🚨 BREAKING: Someone built a coding editor that le...
🚨 BREAKING: Someone built a coding editor that lets you use Claude Opus 4.6, GPT-5.4 and Gemini 3.1 Pro without paying a single dollar in API fees. It's called Glass and it just made Cursor, Windsurf, and every AI IDE look like a rip-off. Here's how it works (in plain English):↓
Gemma 4 is here! Our most intelligent open models...
Gemma 4 is here! Our most intelligent open models to date, are built on the same world-class research and tech as Gemini 3, and are sized to run and fine-tune efficiently on local hardware. Check out what @GoogleGemma 4 brings to devs: 💎 Advanced Reasoning: Deep logic tasks, complex multi-step planning, and beyond 💎 Longer context: Seamlessly analyze entire codebases with context windows of 128K tokens for our edge models and 256K tokens for our largest models 💎 Vision and audio: Rich, multimodal interactions out of the box 💎 140+ languages: Trained on 140+ languages 💎 Apache 2.0 license: industry-standard open-source license
"Open source can't compete with frontier AI" Goog...
"Open source can't compete with frontier AI" Google just killed that narrative. Gemma 4 dropped this morning. Gemini 3's architecture. Fully open. Free forever: - 4 sizes ($35 Pi to workstations) - 256K context - Native multimodal - Built-in agent tools - 140+ languages - Beats models 20x larger Apache 2.0. Build anything. Sell it. No permission. No API costs. Data stays local. The future isn't behind corporate APIs. It's on your hardware. Below: Gemma 4 + your own Claude Code setup 👇
let me explain what Google just did: → they’ve j...
let me explain what Google just did: → they’ve just released their most capable open models yet → Gemma 4… built from the same research behind Gemini 3… four sizes… all running on your own hardware → the 31B dense model and 26B mixture of experts model deliver what Google is calling “frontier-level intelligence” on a personal computer... no cloud required… your data stays on your machine → the 26B MoE only activates 3.8B parameters at a time… meaning it runs fast without needing massive compute → the 2B and 4B models are built for phones and edge devices… text, image, and audio support they can see and hear in real time… 140+ languages natively → 256K context window on the larger models… enough to analyze full codebases or handle long multi-turn agent workflows → native tool use built in… these models can plan steps and call tools on your behalf without extra wiring → Arena Elo scores: Gemma 4 31B hit 1464 and the 26B hit 1453… competing with models 20-30x their size as of today… GLM 5 at 754B scored 1469… Kimi k2.5 at 1100B scored 1464… Gemma is doing this at a fraction of the parameters → Apache 2.0 license… fully open weights, commercially permissive… and the first time Google has done this with Gemma → 400 million downloads and over 100,000 community variants since the first Gemma launched → available now on Google AI Studio, HuggingFace, Kaggle, and Ollama the open source AI race just took a massive leap forward imo running frontier-level reasoning on your laptop without sending a single byte to the cloud completely changes the game for privacy, speed, and cost and the fact that a 26B model with 3.8B active parameters is competing with models 20-30x its size tells you where this is heading running models locally? you gotta get this set up today
Gemma 4 is here! 4⃣Our most capable, agentic open...
Gemma 4 is here! 4⃣Our most capable, agentic open model, built on the same research as Gemini 3. ✨ Reasoning. Multimodal. Four sizes (2B to 31B). Base + Instruct. Released under Apache 2.0. Runs on your phone, laptop, or servers. 🧵↓ https://t.co/AMyJQdsljJ
🚨 BREAKING: Every top AI just failed one of the h...
🚨 BREAKING: Every top AI just failed one of the hardest tests ever built. Humans: 100% AI models: nearly 0% • Gemini 3.1 Pro → 0.37% • GPT-5.4 → 0.26% • Opus 4.6 → 0.25% • Grok-4.20 → 0.00% François Chollet just released ARC-AGI-3 — and it’s brutal. 135 new environments. No instructions. No hints. No defined goals. You’re dropped in... and must figure out what “winning” even means. Humans solved everything. AI didn’t cross 1%. Why it matters: → puzzle-like environments → real-time rule discovery → brute force gets punished If humans solve in 10 steps and AI takes 100… it gets 1% - not 10% More compute won’t fix this. Context: ARC-AGI-1 → nearly solved ARC-AGI-2 → massive jump in progress ARC-AGI-3? Reset everything. Launched with Sam Altman at a YC fireside. $2M prize pool. Open-source solutions required. This isn’t a small gap. It shows how far true AGI still is 🚀
Today, we’re launching Gemma 4, our most intellige...
Today, we’re launching Gemma 4, our most intelligent open models to date. Built with the same breakthrough technology as Gemini 3, Gemma 4 brings advanced reasoning to your personal hardware and devices. Here’s what Gemma 4 unlocks for developers: — Intelligence-per-parameter: Our 31B (Dense) and 26B (MoE) models deliver state-of-the-art performance for their size, outcompeting models 20x their size on @arena — Commercial flexibility: Released under a permissive Apache 2.0 license for complete developer flexibility and digital sovereignty — Agentic workflows: Native support for function-calling and structured JSON output allows you to build reliable, autonomous agents — Multimodal edge AI: The E2B and E4B models bring native vision, audio, and low latency to mobile and IoT devices — Long-context reasoning: Up to 256K context windows allow you to process entire repositories or large documents in a single prompt Whether you're building global applications in 140+ languages or local-first AI code assistants, Gemma 4 is built to be your foundation. Explore in @GoogleAIStudio or download the weights on @HuggingFace, @Kaggle, and @Ollama.
Google just dropped Gemma 4, their most powerful o...
Google just dropped Gemma 4, their most powerful open models yet. Built on Gemini 3 tech, now running advanced reasoning and agents directly on your hardware. Fully open under Apache 2.0. https://t.co/ANoCUCrJQV
We just released Gemma 4 — our most intelligent o...
We just released Gemma 4 — our most intelligent open models to date. Built from the same world-class research as Gemini 3, Gemma 4 brings breakthrough intelligence directly to your own hardware for advanced reasoning and agentic workflows. Released under a commercially permissive Apache 2.0 license so anyone can build powerful AI tools. 🧵↓