正在加载视频...

视频加载失败

返回首页

We are in an insane run of open-weight drops. Every modality, open source is winning. This is what an open source AI summer ☀️ looks like: 🧠 LLMs & Reasoning → DeepSeek-V4-Flash-0731 (my king 👑): 304B MoE refresh, Terminal-Bench 2.1 jumps 61.8→82.7 over the preview, DeepSWE 7.3→54.4. Closes in...

victormustar's profile picture

Victor M

29,814 subscribers

54,264 次观看 • 16 天前 •via X (Twitter)

科学技术教育新闻政治
LIVE

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

🎉 Congrats to MiniMax (official) on releasing the open weights for MiniMax H3! Day-0 support in vLLM-Omni! One model reads text, images, video, and audio as a single context and returns video with native stereo audio. Text-to-video, first/last-frame, and multi-reference generation, 4 to 15 seconds at up to 2K and 24 FPS. The MP4 comes back with H.264 video and a synchronized stereo track already muxed in. It serves over the OpenAI-compatible /v1/videos endpoint, synchronous or async with job polling. 🔊 The video ⬇️ was made with H3, served on vLLM-Omni.
0:15

🎉 Congrats to MiniMax (official) on releasing the open weights for MiniMax H3! Day-0 support in vLLM-Omni! One model reads text, images, video, and audio as a single context and returns video with native stereo audio. Text-to-video, first/last-frame, and multi-reference generation, 4 to 15 seconds at up to 2K and 24 FPS. The MP4 comes back with H.264 video and a synchronized stereo track already muxed in. It serves over the OpenAI-compatible /v1/videos endpoint, synchronous or async with job polling. 🔊 The video ⬇️ was made with H3, served on vLLM-Omni.

vLLM

129,674 次观看 • 26 天前

MiniMax H3 from is now available on Venice. Open-weight multimodal video model with native stereo audio on every generation. Available anonymously.
0:20

MiniMax H3 from is now available on Venice. Open-weight multimodal video model with native stereo audio on every generation. Available anonymously.

Venice

17,877 次观看 • 29 天前

167 tok/s on a single RTX 4090. FreeToken just pushed Qwen3.6-35B-A3B NVFP4 into ridiculous territory. → RTX 4090 24GB → 35B total parameters → ~3B active per token → NVFP4 → 167 tok/s decode → 23.55GB peak VRAM → No speculative decoding → No MTP → No draft model That last part is what makes this interesting. The speed isn’t coming from guessing future tokens. FreeToken is exploiting the MoE architecture directly, keeping the active experts on the GPU while using bandwidth-aware execution for the rest. A 35B model with only ~3B active parameters running at 167 tok/s on a consumer 4090 is exactly why MoE models are becoming so compelling for local AI.
0:13

167 tok/s on a single RTX 4090. FreeToken just pushed Qwen3.6-35B-A3B NVFP4 into ridiculous territory. → RTX 4090 24GB → 35B total parameters → ~3B active per token → NVFP4 → 167 tok/s decode → 23.55GB peak VRAM → No speculative decoding → No MTP → No draft model That last part is what makes this interesting. The speed isn’t coming from guessing future tokens. FreeToken is exploiting the MoE architecture directly, keeping the active experts on the GPU while using bandwidth-aware execution for the rest. A 35B model with only ~3B active parameters running at 167 tok/s on a consumer 4090 is exactly why MoE models are becoming so compelling for local AI.

FHILY👑

33,058 次观看 • 3 天前

MiniMax-H3 open source: a general-purpose multimodal generative model that unified understanding of multimodalof text, images, video, and audio, and is capable of generating audio and video with native stereo sound, supporting up to 15 seconds at 2K resolution.
0:06

MiniMax-H3 open source: a general-purpose multimodal generative model that unified understanding of multimodalof text, images, video, and audio, and is capable of generating audio and video with native stereo sound, supporting up to 15 seconds at 2K resolution.

青龍聖者

59,726 次观看 • 29 天前

world models have their comfyui now. introducing alakazam studio, the first local, open source UI for world models. •⁠ ⁠sentence goes in => you edit the state graph => play •⁠ ⁠on a live model, yours, running in your browser •⁠ ⁠⁠apache-2.0, fully open source fork it and point it at your model 👇
0:19

world models have their comfyui now. introducing alakazam studio, the first local, open source UI for world models. •⁠ ⁠sentence goes in => you edit the state graph => play •⁠ ⁠on a live model, yours, running in your browser •⁠ ⁠⁠apache-2.0, fully open source fork it and point it at your model 👇

hugo

25,362 次观看 • 16 天前

Happy Friday! We just put DeepSeek-V4-Pro up on It’s the world’s largest open source model at 1.6T parameters, and you can run it for free running on NVIDIA Blackwell GPUs. Try the NVIDIA NIM API →
0:18

Happy Friday! We just put DeepSeek-V4-Pro up on It’s the world’s largest open source model at 1.6T parameters, and you can run it for free running on NVIDIA Blackwell GPUs. Try the NVIDIA NIM API →

NVIDIA AI

206,000 次观看 • 4 个月前

it's open source time, with a real leap for world models 🎉 NVIDIA's SANA-WM: a camera-conditioned world model that fits on one GPU. 60s of 720p in 34s on a single 5090 - 2.6B params and Apache 2.0!
0:20

it's open source time, with a real leap for world models 🎉 NVIDIA's SANA-WM: a camera-conditioned world model that fits on one GPU. 60s of 720p in 34s on a single 5090 - 2.6B params and Apache 2.0!

Victor M

34,386 次观看 • 3 个月前

The new open-weights Hailuo AI (MiniMax) H3 is live on OpenRouter thanks to MiniMax (official)! H3 is a lightweight video model for text, image, audio, & video input, with native audio-visual output. Built for instruction-led edits, text/brand rendering, and video-2-video motion transfer.
0:15

The new open-weights Hailuo AI (MiniMax) H3 is live on OpenRouter thanks to MiniMax (official)! H3 is a lightweight video model for text, image, audio, & video input, with native audio-visual output. Built for instruction-led edits, text/brand rendering, and video-2-video motion transfer.

OpenRouter

22,767 次观看 • 29 天前

MiniMax H3 is now available on GMI Cloud the #1 open model in Video Arena: across both Text-to-Video and Image-to-Video (video made by MiniMax (official) H3)
0:17

MiniMax H3 is now available on GMI Cloud the #1 open model in Video Arena: across both Text-to-Video and Image-to-Video (video made by MiniMax (official) H3)

GMI Cloud

16,935 次观看 • 24 天前

Gemma 4 12B QAT (dense) achieves 1000+ tokens/sec prefill on 8GB VRAM with 120k context Gemma 4 12B QAT (dense), TurboQuant (Without MTP), RTX 4060 8GB VRAM: Prefill: 1000+ tok/s (42% increase) Decode: 25+ tok/s (25% increase) Context: 120k (150% increase) prefill was 700 tok/sec and decode 20 tok/sec with only 48k context without turbo quant (older test with mtp link in the comments) llama.cpp TurboQuant flags: -m gemma-4-12B-it-qat-UD-Q4_K_XL.gguf -c 120000 --cache-type-k q8_0 --cache-type-v turbo3 -ngl 99 --port 8080 tested with a 27k prompt, 120k context loaded. -ngl 99 here isn't a typo, full 12B dense, every layer on GPU, on an 8GB card. that's the part worth sitting with. The model has vision, audio input, thinking/reasoning and fits your 8GB card. TurboQuant's KV cache savings are what free up the room to do that at 120k context. side by side with yesterday: 26B A4B MoE got 320+ tok/s prefill. this dense 12B is clearing 1000+ rig: RTX 4060 8GB · i7H · 16GB RAM same two flags as yesterday, different model size: --cache-type-k q8_0 --cache-type-v turbo3 thanks to TheTom/llama-cpp-turboquant, TurboQuant fork of llama.cpp by Tom Turney (Tom Turney) to make this work. unsloth's model quant huggingface and the llama.cpp fork github link in the comments Do you prefer a dense or a MoE for your 8GB card?
0:12

Gemma 4 12B QAT (dense) achieves 1000+ tokens/sec prefill on 8GB VRAM with 120k context Gemma 4 12B QAT (dense), TurboQuant (Without MTP), RTX 4060 8GB VRAM: Prefill: 1000+ tok/s (42% increase) Decode: 25+ tok/s (25% increase) Context: 120k (150% increase) prefill was 700 tok/sec and decode 20 tok/sec with only 48k context without turbo quant (older test with mtp link in the comments) llama.cpp TurboQuant flags: -m gemma-4-12B-it-qat-UD-Q4_K_XL.gguf -c 120000 --cache-type-k q8_0 --cache-type-v turbo3 -ngl 99 --port 8080 tested with a 27k prompt, 120k context loaded. -ngl 99 here isn't a typo, full 12B dense, every layer on GPU, on an 8GB card. that's the part worth sitting with. The model has vision, audio input, thinking/reasoning and fits your 8GB card. TurboQuant's KV cache savings are what free up the room to do that at 120k context. side by side with yesterday: 26B A4B MoE got 320+ tok/s prefill. this dense 12B is clearing 1000+ rig: RTX 4060 8GB · i7H · 16GB RAM same two flags as yesterday, different model size: --cache-type-k q8_0 --cache-type-v turbo3 thanks to TheTom/llama-cpp-turboquant, TurboQuant fork of llama.cpp by Tom Turney (Tom Turney) to make this work. unsloth's model quant huggingface and the llama.cpp fork github link in the comments Do you prefer a dense or a MoE for your 8GB card?

Alok

34,500 次观看 • 2 个月前

Topview is one of the first platforms on earth to bring MiniMax H3 to creators ,in partnership with Hailuo AI (MiniMax). Native 2K. 15-second clips. Multimodal understanding across text, image, audio, video. And it's 30% of Seedance 2.0's price. 🔥 TopviewAI
0:15

Topview is one of the first platforms on earth to bring MiniMax H3 to creators ,in partnership with Hailuo AI (MiniMax). Native 2K. 15-second clips. Multimodal understanding across text, image, audio, video. And it's 30% of Seedance 2.0's price. 🔥 TopviewAI

LudovicCreator

211,570 次观看 • 29 天前

Another demo of the iPhone 17 Pro’s on-device LLM performance This time with Ling mini 2.0 by InclusionAI, a 16B MoE model with 1.4B active parameters running at ~120tk/s Thanks to Awni Hannun for the MLX DWQ 2-bit quants
0:11

Another demo of the iPhone 17 Pro’s on-device LLM performance This time with Ling mini 2.0 by InclusionAI, a 16B MoE model with 1.4B active parameters running at ~120tk/s Thanks to Awni Hannun for the MLX DWQ 2-bit quants

Adrien Grondin

46,205 次观看 • 11 个月前

MiniMax H3 is built for native multimodal understanding, combining text, images, videos, and audio in one context. Creators can use diverse references beyond text prompts, with better control over style, motion, sound, and characters. As an open-weight model, H3 unlocks new possibilities for AI video creation and editing
0:15

MiniMax H3 is built for native multimodal understanding, combining text, images, videos, and audio in one context. Creators can use diverse references beyond text prompts, with better control over style, motion, sound, and characters. As an open-weight model, H3 unlocks new possibilities for AI video creation and editing

Tyler Wayne

80,546 次观看 • 25 天前

MiniMax H3 is now live on TopviewAI. Native 2K output and 15 second clips are useful on their own, but the real advantage is the ability to work across text, image, audio, and video in one workflow. At only 30% of Seedance 2.0’s price, this is worth testing.
0:15

MiniMax H3 is now live on TopviewAI. Native 2K output and 15 second clips are useful on their own, but the real advantage is the ability to work across text, image, audio, and video in one workflow. At only 30% of Seedance 2.0’s price, this is worth testing.

Doreen

21,862 次观看 • 29 天前

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 Google Gemma 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
0:19

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 Google Gemma 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

Google for Developers

269,912 次观看 • 4 个月前

This AI model turns a single image + audio into a talking avatar video of infinite length. StableAvatar is built on Wan 2.1, and MIT Licensed. Open-source AI video is accelerating, and it’s exciting to watch!
0:19

This AI model turns a single image + audio into a talking avatar video of infinite length. StableAvatar is built on Wan 2.1, and MIT Licensed. Open-source AI video is accelerating, and it’s exciting to watch!

Miguel | AP

35,804 次观看 • 1 年前

🎉 Congrats to Thinking Machines on TML Inkling—a 1T-parameter open-weight model supported in vLLM from Day 0. Highlights: • Natively multimodal across text, image, and audio • Up to 1M-token context • New architecture with relative attention, short convolutions, and MoE expert sinks • 8 MTP heads for speculative decoding vLLM supports both NVFP4 and BF16 checkpoints, optimized for NVIDIA Blackwell and Hopper, reaching up to 380 tok/s/user on 4× GB200 with MTP. Huge thanks to the Thinking Machines Lab team for the close collaboration 🙏 Read about implementation below 👇
0:16

🎉 Congrats to Thinking Machines on TML Inkling—a 1T-parameter open-weight model supported in vLLM from Day 0. Highlights: • Natively multimodal across text, image, and audio • Up to 1M-token context • New architecture with relative attention, short convolutions, and MoE expert sinks • 8 MTP heads for speculative decoding vLLM supports both NVFP4 and BF16 checkpoints, optimized for NVIDIA Blackwell and Hopper, reaching up to 380 tok/s/user on 4× GB200 with MTP. Huge thanks to the Thinking Machines Lab team for the close collaboration 🙏 Read about implementation below 👇

vLLM

49,212 次观看 • 1 个月前

LongCat-2.0 from Meituan LongCat is free in Nous Portal for one week. It is a 1.6T-parameter MoE with 1M context built for agentic coding, scores 70.8 on Terminal-Bench 2.1, and can ingest an entire codebase in one pass. Try it out at
0:19

LongCat-2.0 from Meituan LongCat is free in Nous Portal for one week. It is a 1.6T-parameter MoE with 1M context built for agentic coding, scores 70.8 on Terminal-Bench 2.1, and can ingest an entire codebase in one pass. Try it out at

Nous Research

99,022 次观看 • 15 天前

Liquid's LFM2.5-8B-A1B smashed OpenAI's gpt-oss-20b on tool calling We ran both locally on a MacBook Pro M5 Max, 64GB, and gave each the same trip-planning request that only completes if the model fires all 7 tool calls - weather for 3 cities, two currency conversions, an email and a reminder Outputs: LFM2.5-8B-A1B: 4.8 GB RAM usage, 7/7 tool-calls, 266 tok/s, 6.9s OpenAI gpt-oss-20b: 11 GB RAM usage, 3/7 tool-calls, 146 tok/s, 15.0s The 8B used less than half the RAM and still fired all 7 calls, while the 20B silently dropped more than half of its own. It also ran ~2x faster, wrapping the full agentic request in 6.9s against 15s. That's what 38T training tokens buy: a 1B-active MoE that nails the agentic tool calls a model 2.5x its active size keeps dropping
0:15

Liquid's LFM2.5-8B-A1B smashed OpenAI's gpt-oss-20b on tool calling We ran both locally on a MacBook Pro M5 Max, 64GB, and gave each the same trip-planning request that only completes if the model fires all 7 tool calls - weather for 3 cities, two currency conversions, an email and a reminder Outputs: LFM2.5-8B-A1B: 4.8 GB RAM usage, 7/7 tool-calls, 266 tok/s, 6.9s OpenAI gpt-oss-20b: 11 GB RAM usage, 3/7 tool-calls, 146 tok/s, 15.0s The 8B used less than half the RAM and still fired all 7 calls, while the 20B silently dropped more than half of its own. It also ran ~2x faster, wrapping the full agentic request in 6.9s against 15s. That's what 38T training tokens buy: a 1B-active MoE that nails the agentic tool calls a model 2.5x its active size keeps dropping

atomic.chat

90,063 次观看 • 3 个月前