What are the hardware requirements for Soren 1 Small?

Soren 1 Small requires approximately 3.8 GB of VRAM and supports a context window of 1024K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Soren 1 Small and how large is it?

Soren 1 Small is developed by syntropy-ai, a model with 2b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Soren 1 Small?

Soren 1 Small is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Soren 1 Small by syntropy-ai — VRAM 3.8GB, 1024K context

Name: Soren 1 Small
Author: syntropy-ai

Soren 1 Small is an open-source language model by syntropy-ai. Features: 2b LLM, VRAM: 3.8GB, Context: 1024K, License: apache-2.0, Instruction-Based, Merged, LLM Explorer Score: 0.26.

Merged Model 1048576-context 1m-context Agentic Analytical Assistant Base model:adapter:qwen/qwen3.... Base model:qwen/qwen3.5-2b Blackwell Chat Claude-style Code-assistant Code-generation Coding Coding-model Compact-model Conversational Dataset:angrygiraffe/claude-op... Dataset:argilla/distilabel-cap... Dataset:argilla/openhermespref... Dataset:code-refinement/dpo-sa... Dataset:dalisoft/claude-opus-4... Dataset:hastagaras/claude-sonn... Dataset:ianncity/hunter-alpha-... Dataset:m-a-p/codefeedback-fil... Dataset:nettoov/gpt-5.4-xhigh-... Dataset:nvidia/helpsteer Dataset:openbmb/ultrafeedback Dataset:roman1111111/claude-op... Dataset:roman1111111/claude-so... Dataset:sashvat/hyperthink-x-n... Dataset:syntropy-ai/its-me-sor... Dataset:teichai/claude-4.5-opu... Dataset:teichai/claude-opus-4.... Dataset:teichai/claude-sonnet-... Dataset:teichai/gpt-5-codex-25... Dataset:teichai/gpt-5.2-high-r... Dataset:teichai/lordx64-claude... Dataset:vezora/code-preference... Decoder-only Dpo Edge-ai En Helpful-assistant Honest-ai Human-like Instruct Instruction-following Local-llm Long-context Long-context-model Lora Low-hallucination Math Nvidia Open-source-ai Open-weights Preference-optimization Problem-solving Project-syntropic Python Qwen Qwen3 Qwen3.5 Qwen3.5-2b Qwen3 5 text Reasoning Reasoning-assistant Reasoning-model Region:us Rtx-pro-6000-blackwell Safetensors Sft Software-engineering Soren Soren-1 Soren-1-small Supervised-finetuning Syntropy-ai Thinking Trl Unsloth Yarn Yarn-4x

Model Card on HF 🤗: https://huggingface.co/syntropy-ai/Soren-1-Small

Soren 1 Small Parameters and Internals

LLM Name	Soren 1 Small
Repository 🤗	https://huggingface.co/syntropy-ai/Soren-1-Small
Model Name	soren_dpo2_merged
Base Model(s)	Qwen/Qwen3.5-2B Qwen/Qwen3.5-2B
Merged Model	Yes
Model Size	2b
Required VRAM	3.8 GB
Updated	2026-07-31
Maintainer	syntropy-ai
Model Type	qwen3_5_text
Instruction-Based	Yes
Model Files	3.8 GB
Supported Languages	en
Model Architecture	Qwen3_5ForCausalLM
License	apache-2.0
Context Length	1048576
Model Max Length	1048576
Transformers Version	5.9.0
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	248320
Errors	replace

Rank the Soren 1 Small capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Soren 1 Small

Best Alternatives	Context / RAM	Downloads	Likes
Noema 1.5 2B	256K / 3.8 GB	322	0
Noema 2B	256K / 3.8 GB	385	1

Note: green Score (e.g. "73.2") means that the model is better than syntropy-ai/Soren-1-Small.