Powering the future of sovereign AI across Australia (https://t.co/KPsqhwYpxz), Europe (@InfercomAI) & the UK (Argyll) ⚡️
Our AI infrastructure is helping nations scale AI securely, responsibly, & sustainably, partnering with leaders to tackle the data center power crisisPowering the future of sovereign AI across Australia (https://t.co/KPsqhwYpxz), Europe (@InfercomAI) & the UK (Argyll) ⚡️
Our AI infrastructure is helping nations scale AI securely, responsibly, & sustainably, partnering with leaders to tackle the data center power crisisPowering the future of sovereign AI across Australia (https://t.co/KPsqhwYpxz), Europe (@InfercomAI) & the UK (Argyll) ⚡️
Our AI infrastructure is helping nations scale AI securely, responsibly, & sustainably, partnering with leaders to tackle the data center power crisisPowering the future of sovereign AI across Australia (https://t.co/KPsqhwYpxz), Europe (@InfercomAI) & the UK (Argyll) ⚡️
Our AI infrastructure is helping nations scale AI securely, responsibly, & sustainably, partnering with leaders to tackle the data center power crisisPowering the future of sovereign AI across Australia (https://t.co/KPsqhwYpxz), Europe (@InfercomAI) & the UK (Argyll) ⚡️
Our AI infrastructure is helping nations scale AI securely, responsibly, & sustainably, partnering with leaders to tackle the data center power crisis.
📣 @RodrigoLiang just landed on @FastCompany's AI 20 for 2025, a list celebrating the leaders pushing AI in unexpected directions.
Huge thanks to our customers, partners, and team for powering this journey.
https://t.co/YusUvNLB48
🚀 @Stanford research just introduced "Intelligence per Joule", the ultimate metric for AI efficiency. And our RDU delivers 4X more intelligence per joule Blackwell. ⚡️
🔗 Dive into the research in our blog: https://t.co/Yl6YIjN4zj
🚀 @OVHcloud has chosen SambaNova to power its new ultra-low-latency, high-scale AI Endpoints.
With SambaStack, OVH delivers:
• Fast inference
• Energy efficiency
• 99.8% uptime SLA
• Largest open-source models support
• Real-time & batch endpoints
https://t.co/0VP4E7X4Bv
It's here. 🚀 In partnership with @Stanford & @UCBerkeley, we're releasing Agentic Context Engineering (ACE) on GitHub, your complete framework for building self-improving AI agents.
The future of agentic AI is now open source. Dive into the blog & code: https://t.co/777XkOLeqE
Infercom is leading the sovereign AI transformation in Europe 🚀
Altug Eker, Managing Director @InfercomAI, shares his perspective on why sovereign AI is critical for businesses and governments in Europe.
https://t.co/BU4Z7C0h4s
Meet the SambaNova SDK 🚀
Build AI apps faster with native, type-safe SDKs for Python & JavaScript/TypeScript.
Access top open-source models like @DeepSeek_ai, @AIatMeta Llama & @Alibaba_Qwen on SambaCloud and get started in minutes.
Read more: https://t.co/FdscFO8h3g
We’re powering the future of sovereign AI—across Australia, Europe & the UK.
Our energy-efficient AI infrastructure is tackling the data center power crisis and enabling nations to scale AI securely, responsibly, and sustainably.
Full release:
State of DeepSeek models: DeepSeek launched its new V3.1 Terminus and V3.2 Exp hybrid reasoning models in quick succession in September, making meaningful steps in both intelligence and cost efficiency
Both models can be used in reasoning and non-reasoning modes, and are intended to replace both the prior V3 and R1 models. DeepSeek V3.1 Terminus jumped 4 points higher on the Artificial Analysis Intelligence Index than the original V3.1 release without substantially more tokens, and V3.2 Experimental retained most of these gains while substantially reducing per-token costs due to DeepSeek’s experimental new architecture.
DeepSeek V3.1 Terminus and V3.2 Exp score similarly in Artificial Analysis, with V3.1 Terminus coming in one point above V3.2 Exp. Both are material upgrades to earlier V3 and R1 models.
A range of providers are serving both models, with availability of V3.2 still expanding:
➤ DeepSeek V3.1 Terminus is available on SambaNova, DeepInfra, Fireworks, GMI, and Novita, with @SambaNovaAI offering output speeds of ~250 tokens per second (around 10x faster than DeepSeek first-party inference)
➤ DeepSeek V3.2 Exp is available on DeepSeek’s first-party API, DeepInfra, GMI, and Novita - with DeepInfra serving it fastest at speeds of up to 79 tokens per second
AI inference is a growing bottleneck & it’s hitting data centers hard.
Listen to the latest ep of @theneurondaily with Kwasi Ankomah, @CoreyNoles & Grant Harvey as they discuss inference throughput, power efficiency & the future of AI infrastructure.
📣 @hume_ai's EVI 4 Mini (Empathic Voice Interface) & Octave 2 EXCLUSIVELY available on SambaCloud — real-time expressive AI, fluent in Japanese + 10 languages.
-100-300ms ultra-low latency
-Human-like emotional intelligence
-Enterprise security + scalability
We're part of the @FortuneMagazine's Future50 2025 list, that spotlights companies defining tomorrow’s economy
For the 2nd year in a row 🏆
Honored to be recognized with the world’s most innovative growth leaders. This fuels our mission to push the boundaries of enterprise AI
🐋 @DeepSeek_ai-R1-0528 just leveled up.
Now with function calling + reasoning on par with OpenAI & Gemini. And it's all open-source so you see the reasoning.
Check the benchmarks below.