Model Gallery

Discover and install AI models from our curated collection

2 models available
1 repositories
Documentation

Find Your Perfect Model

Filter by Model Type

Browse by Tags

Attention: Trust Remote Code is required for this model
phonellm-alpha-1:vllm
PhoneLLM Alpha 1 is a 30B-parameter, 3.5B-active mixture-of-experts model tuned by the Pipecat team for low-latency voice agents, multi-turn conversations, and accurate tool use without thinking. This entry serves the bfloat16 safetensors with vLLM and supports a 262K-token context. PhoneLLM is released under BSD-2-Clause and derives from NVIDIA Nemotron 3 Nano 30B-A3B. The underlying work remains subject to the NVIDIA Nemotron Open Model License and its redistribution and attribution requirements.

Repository: localaiLicense: bsd-2-clause

Attention: Trust Remote Code is required for this model
phonellm-alpha-1:sglang
PhoneLLM Alpha 1 is a 30B-parameter, 3.5B-active mixture-of-experts model tuned by the Pipecat team for low-latency voice agents, multi-turn conversations, and accurate tool use without thinking. This entry serves the bfloat16 safetensors with SGLang and supports a 262K-token context. PhoneLLM is released under BSD-2-Clause and derives from NVIDIA Nemotron 3 Nano 30B-A3B. The underlying work remains subject to the NVIDIA Nemotron Open Model License and its redistribution and attribution requirements.

Repository: localaiLicense: bsd-2-clause