Models & Labs
Baseten Joins Hugging Face Inference Providers
Hugging Face Bloghigh confidence
Why it matters
- →Baseten's integration expands the range of AI models available on Hugging Face, enhancing developer options.
- →The flexibility in billing and API key usage offers developers more control over their AI infrastructure costs.
- →This collaboration simplifies the process of integrating AI capabilities into applications, reducing setup time.
Baseten has become a supported Inference Provider on the Hugging Face Hub, enhancing serverless AI capabilities for developers. This integration allows developers to access a variety of models, including popular LLMs like Kimi K3 and DeepSeek V4 Flash, directly through Hugging Face's platform. Users can choose to route requests via Hugging Face or use their own API keys, providing flexibility in billing. The integration is available through Hugging Face's SDKs for Python and JavaScript, making it easier for developers to incorporate AI into their applications.
Read originalThe story around this
TopicHugging Face Partnerships And ToolsCooling
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
More from Hugging Face Blog
© Hugging Face BlogModels & Labsother
Liquid AI releases DSpark drafter for vision models
Liquid AI is tackling the latency bottleneck in vision-language models with a specialized speculative decoding drafter. The LFM2.5-VL-DSpark model adds only 280M parameters to the base 3B architecture but delivers up to 3.13x faster on-device decoding and 2.66x speedups on H100 GPUs. By integrating directly with llama.cpp, MLX-VLM, and SGLang from day one, they are making high-performance VLM inference accessible without complex custom infrastructure. This shifts the focus from raw model size to efficient token drafting for edge deployment.
Hugging Face Blog
© Hugging Face BlogMore in Models & Labs
Models & Labsother
llama.cpp b11193 adds Snapdragon Hexagon and ROCm 10
This release quietly expands llama.cpp's hardware support to include Qualcomm's Hexagon NPU on Linux arm64, a significant step for local inference on Snapdragon devices. It also updates CUDA builds to version 13.4 and introduces ROCm 10.0 binaries, keeping the project aligned with the latest NVIDIA and AMD driver ecosystems. KleidiAI on Apple Silicon is temporarily disabled in this build, likely due to stability checks rather than a feature rollback. For developers targeting edge AI or diverse GPU stacks, this update ensures broader compatibility without requiring custom compilation.
llama.cpp Releases
© Lev SelectorModels & Labsmodels
NVIDIA NVFP4 and SoL-Pi Token Efficiency
NVIDIA introduced the NVFP4 4-bit format and SoL-Pi technology, which uses 2x fewer tokens for improved efficiency.