Skip to content
Weights Not Walls
Menu

sonar

by dphnAI

In its repository's words: “Large-scale LLM inference engine”

Repository on GitHubsonar.dphn.ai

Facts from GitHub

Licence
AGPL-3.0 (GNU Affero General Public License v3.0), copyleft, our reading
Activity
Last push 17 Nov 2026: pushed to in the last 90 days
Latest release
not fetched yet
Stars
1.9k
Main language
Python
On GitHub since
23 Jun 2023
Topics
api-rest, cuda, inference-engine, inferentia, intel, lora, machine-learning, rocm, speculative-decoding, tpu

What it says it supports

Loads
safetensors, FP8
Runs on
Linux
Hardware
NVIDIA, AMD

Each item is on our list for this tool and named in its README. A mention is not a test: check the project's documentation for your setup.

Reviewed models it can load

Matched by file format only, from our snapshot of the Hub: GGUF where we measured GGUF files, other formats where a conversion exists, safetensors for the original weights. Support for a new architecture can lag behind its release.

safetensors 428

openai-community/gpt2, microsoft/Florence-2-base, Qwen/Qwen3-0.6B, Qwen/Qwen3-14B, deepseek-ai/DeepSeek-V3.2, Qwen/Qwen3.8-27B, Qwen/Qwen3.5-9B, Qwen/Qwen2.5-7B-Instruct, zai-org/GLM-5.3-Flash, Qwen/Qwen3-VL-8B-Instruct, google/gemma-4-31B-it, Qwen/Qwen3-8B, and 416 more, most downloaded first

FP8 235

openai-community/gpt2, Qwen/Qwen3-0.6B, Qwen/Qwen3-14B, Qwen/Qwen3.8-27B, Qwen/Qwen3.5-9B, Qwen/Qwen2.5-7B-Instruct, zai-org/GLM-5.3-Flash, Qwen/Qwen3-VL-8B-Instruct, google/gemma-4-31B-it, Qwen/Qwen3-8B, Qwen/Qwen3.5-2B, Qwen/Qwen2.5-VL-7B-Instruct, and 223 more, most downloaded first

Other tools in Serve

All of Serve

Tool data from the GitHub API, fetched 19 Nov 2026, 15:43 UTC. Listing is not endorsement and nobody can pay to be listed. Something wrong here? Write to hello@weightsnotwalls.com.