Alguien creó un repo en GitHub que te permite ejecutar Claude Code completamente gratis. Redirige tu tráfico de Claude Code a 10 proveedores gratuitos como DeepSeek y Kimi, toma 5 minutos configurarlo, y ya tiene a más de 20,000 desarrolladores ejecutándolo.

3.3b tokens served so far added GPUs so we're not bumping into any limits tok / sec - 116.2 tok / min - 70.3 million time to first tok - 0.71s

Out of GPUs. anyone know where I can get a B300 node at a reasonable price? can do 3 year commit asap i want to keep offering free AI

Gemini 4 Argon is here. It’s available globally today for developers and enterprises in the paid tier of the Gemini API and the Gemini platform, for Ultra subscribers in Google Antigravity, and Ultra and Pro subscribers in the Gemini app and Search. Hope you all enjoy it!

probably going to run this until Monday what should I do with the 8xH100 node after? considering free Mimo 2.6 Flash

just rented an 8xH100 node for a few months what should I do with it? considering hosting GLM 5.3 Flash and giving it away for free

Every AI model. One key. Zero markup. That’s the starting point. The bigger ambition: help companies turn their AI traffic into better models they own @silennai helped grow AutoGPT to 160,000 GitHub stars. Now, at @explabsai, he and cofounder @OKfallah are working to bring model access, evaluation and training into one unified experience. The open-source gateway gives companies control over model access. From there, the team is building simulations around their traffic to understand different workloads and recommend models based on three things: cost, quality and speed. The goal goes beyond choosing a model. It’s helping companies evaluate performance and train specialized models around their own work when they need them. Silen expects agents to do most of the work. So the product has to serve both: humans who need to understand what’s happening and agents that need to use it. Start with access. Build toward ownership. "The open source AI gateway. Hosted providers, your own keys, and your own GPUs behind one endpoint, at the provider's price" ‍ 🎙️ @silennai, CTO & Co-Founder, @explabsai on @fondocom @thestartpod 02:51 Going off the beaten path: TKS, Minerva & dropping out 03:37 Getting into YC & the research behind Experiential Labs 05:00 From 8 H100s and continual learning to building a company 05:58 Why every company will become an AI company 06:42 Cost, quality & speed: the “holy trinity” of AI models 07:00 When fine-tuning beats prompting 08:14 What happens when powerful AI gets incredibly cheap 09:00 Why the future could be a model—or LoRA—per customer 10:00 Open-source models get the tokens; closed models get the spend 10:27 Jevons paradox, DeepSeek & exploding AI usage 11:06 Why OpenAI and Anthropic aren’t the next Yahoo 13:05 Experiential Labs’ bigger vision beyond the AI gateway 13:45 Building AI infrastructure for humans and agents 14:01 Optimizing models across cost, quality & speed

MiMo-V2.6-Pro debuts as the top open weights model on the Artificial Analysis Intelligence Index (46). At $0.13 per Intelligence Index task, it lands on the Intelligence vs. Cost per Task Pareto frontier @Xiaomi has just released MiMo-V2.6-Pro, an open weights model with major advances in intelligence over its predecessor, MiMo-V2.5-Pro (Intelligence Index: 26). Despite the improvement, it retains the same attractive pricing at $0.435 per 1M input tokens (with a 99% cache-hit discount) and $0.87 per 1M output tokens. This makes MiMo-V2.6-Pro one of the most cost-efficient models to deploy. MiMo-V2.6-Pro is an MoE model with 1.02T total parameters and 42B active parameters. Stay tuned for additional analysis of the model. Check out MiMo-V2.6-Pro full benchmarking breakdown here: artificialanalysis.ai