Florian S
@airesearch12
Followers
4K
Following
5K
Media
1K
Statuses
7K
since age of 6, it's been me & the code. architecting tech for a fortune 500 + a lot of hobby projects (some slightly profitable). currently focusing on ai.
Joined October 2022
Chutes leading in Open Source inference volume on @OpenRouterAI is actually crazy news. And it's still such a small project! Let me explain just *how* small in comparison to its competitors....
so... we just found out something absolutely insane 🧵 we weren't even looking for this data. one of our engineers was going through OpenRouter's Publicly available charts at 3am (as you do) and discovered something that made us all stop and stare at our screens.
2
7
58
Is this model capable for a simple AI referee app?
The model can perform global summarization on long videos while retaining the ability to perform fine-grained reasoning on temporal clues, such as summarizing goal events and timestamps in a full football match.
0
0
0
GLM-4.6V can accept multimodal inputs of various types and automatically generate high-quality, structured image-text interleaved content.
1
4
90
Even if progress were to stall rn, it wouldn't be another AI winter. We've reached spring, once and for all. Countless seeds have been sown and will grow into magnificent plants. The extent of untapped potential is staggering.
Age of Research would be a beautiful name for an AI winter.
0
0
0
Yes
1
0
16
@Zai_org just released GLM-4.6V on @huggingface, amazing model given its size. - Native multimodal function calling!!! - 128k context length - 106A12B and 9B (for flash variant) - native integration with Transformers 5, sglang and vllm - mit license (as always) Great work
10
26
230
Now that's a lot of money "Anthropic signs $52B deal to purchase Google's TPUv7 chips directly."
1
0
4
Congrats, @tngtech! "At this moment, 38x8=304 H200s are processing up to 85k requests per hour, and about 1.5 million requests per day via @chutes_ai"
1
2
8
We’ve been quietly working on the most important upgrade in Chutes history. If you care about security and who sees your data, you’ll want to be here next week. Stay tuned 👀 .
19
52
323
Our new Qwen3-TTS (version 2025-11-27) is here! 🚀 We've leveled up on what matters most: ✨ More Personalities: Over 49 high-quality voices, from cute and playful to wise and stern. Find your perfect match! 🌍 Global Reach: Now speaks 10 languages (zh, en, de, it, pt, es, ja,
70
288
2K
RWKV-7 G0b 13.3B: pure RNN surpassing Qwen3 14B in some evals, without eval-maxxing. And 1B / 3B updated. GGUF updated too.
RWKV-7 G0a3 13.3B: strongest pure RNN ever, MMLU 76.0% (+CoT=82.5%), MATH500 76.0%, GSM8K 92.3%, MMLU Pro 49.8% (+CoT=61.6%), no eval-maxxing / mid-training / post-training. Download: https://t.co/oJxpuJoeQD Ollama: https://t.co/bScaOJSoD0
0
20
142
LM Arena is not where the whale shines most
🚨BREAKING: Text Leaderboard Update 🐳 Deepseek-v3.2 enters the leaderboard at #38, and Deepseek-v3.2-thinking lands at #41. For comparison, previous versions ranked higher: 🔹 v3.2 ranks #38 (–5 pts v3.1 and –14 pts v3.2-exp) 🔹 v3.2-thinking ranks #41 (–7 pts vs v3.1-thinking
0
0
0
🚨GOOGLE ANTIGRAVITY HAS FREE OPUS 4.5 THINKING🚨 Hey guys, you are welcome 😜. Antigravity as a package is insane right now with the inclusion of Opus 4.5. Is OpenAI dead? We will see next week I guess, but they are so so behind😭 Also, I tried GPT-5.1-Codex-Max vs. Opus 4.5
127
167
2K
Will Smith made a fortune selling spaghetti eating training material to ai video gen labs.
Kling 2.6 with Native Audio just dropped on Higgsfield Uncensored, unrestricted video with audio generation in 1080p. 8 wild examples:👇 1. Will Smith eating spaghetti
0
0
0
Kimi K2 Thinikng by @Kimi_Moonshot still more intelligent than te new DeepSeek V3.2. 👀 @crystalsssup
DeepSeek V3.2 is the #2 most intelligent open weights model and also ranks ahead of Grok 4 and Claude Sonnet 4.5 (Thinking) - it takes DeepSeek Sparse Attention out of ‘experimental’ status and couples it with a material boost to intelligence @deepseek_ai V3.2 scores 66 on the
5
2
68
🤔
Why the heck do all these Chinese robots prepare for infantry warfare?!
0
0
2
🚨 It is here! Kling 2.6 is launching exclusively on fal day 0! 🎬 Native audio generation for text-to-video and image-to-video 🎵 Cinematic storytelling with expressive audio performances ✨ High-intensity VFX with detailed sound design
15
36
324
Kling 2.6 now als has video gen with audio
KlingAI released VIDEO 2.6 - Kling AI's First Model with Native Audio It's hard to keep up with all the releases these days, but I'm doing my best. The special thing here is: coherent looking & sounding output, "the 2.6 model opens up narrative possibilities, and makes you "See
0
0
1