Hear from @zsakib_ as he breaks down how to run Wan2.1 and showcases its impressive capabilities
OPEN SOURCE
-
Llama 3.3 Now Supports 128K Context Length on SambaNova
By
–
You asked, we answered 🦾
— SambaNova (@SambaNovaAI) 6 mars 2025
🚀 Our latest update now allows devs to use up to 128K context length on @AIatMeta's Llama 3.3
Experience the full power of Llama 3.3 on SambaNova Cloud 👇#AIYou asked, we answered Our latest update now allows devs to use up to 128K context length on @AIatMeta
's Llama 3.3 Experience the full power of Llama 3.3 on SambaNova Cloud #AI -
Mistral AI étend son portefeuille avec de nouveaux modèles OCR
By
–
Well, it is not OS in fact but available for self-hosting https://
testingcatalog.com/mistral-ai-exp
ands-its-ai-portfolio-with-powerful-new-ocr-models/
… -
Mistral Launches New Free OCR Models
By
–
BREAKING: Mistral AI has released new OCR models – mistral-ocr-2503 and mistral-ocr-latest. Mistral's OCR features are free to try on Le Chat. To test the API, visit La Plateforme.
-
LangChain releases MCP adapters for LangGraph integration
By
–
if you're looking to use langgraph mcp… we released something a week or so ago! https://
changelog.langchain.com/announcements/
mcp-adapters-for-langchain-and-langgraph
… -
AI21 Launches Jamba 1.6: Top Open Model for Enterprise
By
–
Today we launched Jamba 1.6, the best open model for private enterprise deployment. AI21’s Jamba outperforms Cohere, Mistral and Llama on key benchmarks, including Arena Hard, and rivals leading closed models while maintaining unmatched speed and quality. Now available on
-
Alibaba QwQ-32B: Reasoning AI Matches DeepSeek-R1 Cheaper
By
–
Alibaba’s Qwen team dropped QwQ-32B, a reasoning AI that matches or surpasses DeepSeek-R1 at much less cost
— Rowan Cheung (@rowancheung) 6 mars 2025
—20x smaller than DeepSeek-R1
—Priced $0.20 per million input and output tokens
—Open-sourced under Apache 2.0
Here it is running on M4 Max:pic.twitter.com/J40JG0jGtCAlibaba’s Qwen team dropped QwQ-32B, a reasoning AI that matches or surpasses DeepSeek-R1 at much less cost —20x smaller than DeepSeek-R1
—Priced $0.20 per million input and output tokens
—Open-sourced under Apache 2.0 Here it is running on M4 Max: -
Alibaba Qwen 32B Full Release Matches DeepSeek R1 Performance
By
–
We’ve been hyped to support @Alibaba_Qwen 32B during its preview phase & now the full version is here, performing on par with #DeepSeek R1! Devs, get ready—our update drops soon! More info on @huggingface
-
HeadInfer: Long-Context LLM Inference on Consumer GPUs
By
–
HeadInfer: Unlocking Long-Context LLM Inference on Consumer GPUs (Million-level Tokens)
*long-context inputs require large GPU memory.
*A standard LLM like Llama-3–8B requires 207GB of GPU memory for 1 million tokens — far beyond the capabilities of consumer GPUs like the RTX