What are the hardware requirements for TwIL LM3 Pro?

TwIL LM3 Pro requires approximately 2.2 GB of VRAM and supports a context window of 128K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed TwIL LM3 Pro and how large is it?

TwIL LM3 Pro is developed by webAI-Official, a model with 3b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does TwIL LM3 Pro perform on standard benchmarks?

TwIL LM3 Pro has the following published scores: LMSYS ELO 1290. Compare against reference models on this page or on the LLM Explorer leaderboards.

TwIL LM3 Pro by webAI-Official — VRAM 2.2GB, 128K context

Name: TwIL LM3 Pro
Author: webAI-Official

TwIL LM3 Pro is an open-source language model by webAI-Official. Features: 3b LLM, VRAM: 2.2GB, Context: 128K, License: other, Quantized, LLM Explorer Score: 0.47, ELO: 1290.

Base model:adapter:ibm-granite... Base model:ibm-granite/granite... Conversational En Endpoints compatible Formal-logic Gguf Granite Granite-4.2 Grpo Lora Model-merging Q4 Quantized Reasoning Region:us Safetensors Wise-ft

Model Card on HF 🤗: https://huggingface.co/webAI-Official/TwIL-LM3-Pro

TwIL LM3 Pro Benchmarks

LMSys ELO 1290 vs 1272 (so35) 1.4%

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

TwIL LM3 Pro Parameters and Internals

LLM Name	TwIL LM3 Pro
Repository 🤗	https://huggingface.co/webAI-Official/TwIL-LM3-Pro
Base Model(s)	ibm-granite/granite-4.2-3b ibm-granite/granite-4.2-3b
Model Size	3b
Required VRAM	2.2 GB
Updated	2026-10-04
Maintainer	webAI-Official
Model Type	granite
Model Files	7.3 GB 2.2 GB 2.6 GB 3.0 GB 3.9 GB 7.3 GB
Supported Languages	en
GGUF Quantization	Yes
Quantization Type	gguf\|q4\|q4_k\|q5_k
Model Architecture	GraniteForCausalLM
License	other
Context Length	131072
Model Max Length	131072
Transformers Version	5.15.0
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|im_end\|>
Vocabulary Size	100352
Errors	replace

Rank the TwIL LM3 Pro 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 TwIL LM3 Pro

Best Alternatives	Context / RAM	Downloads	Likes
Granite 4.1 3B Fp8	128K / 4.2 GB	2169	8
Granite 4.2 3B MLX 4bit	128K / 2.1 GB	132	1
Granite 4.2 3B	128K / 7.4 GB	1894	33
Granite 4.1 3B	128K / 6.8 GB	328863	100
Granite 4.2 3B Heretic	128K / 7.4 GB	626	1
Granite 4.2 3B	128K / 7.4 GB	300	1
Granite 4.1 3B Base	128K / 6.8 GB	8994	21
...e V1 Clm Nasasmd Granite4.1 3B	128K / 6.9 GB	203	1
...anite 4.1 3B Medical Reasoning	128K / 6.8 GB	262	1
... Granite 4.1 3B Claude Fable 5	128K / 6.8 GB	519	1