What are the hardware requirements for Darwin 27B RSI?

Darwin 27B RSI requires approximately 53.8 GB of VRAM and supports a context window of 256K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Darwin 27B RSI and how large is it?

Darwin 27B RSI is developed by FINAL-Bench, a model with 27b 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 Darwin 27B RSI?

Darwin 27B RSI is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Darwin 27B RSI by FINAL-Bench — VRAM 53.8GB, 256K context

Name: Darwin 27B RSI
Author: FINAL-Bench

Darwin 27B RSI is an open-source language model by FINAL-Bench. Features: 27b LLM, VRAM: 53.8GB, Context: 256K, License: apache-2.0, LLM Explorer Score: 0.38.

27b Base model:final-bench/darwin-... Base model:finetune:final-benc... Chain-of-thought Conversational Darwin Darwin-rsi Decision-index En Endpoints compatible Final-bench Gpqa Jev Ko Label-free Multilingual No-human-labels Qwen3.5 Qwen3 5 text Reasoning Recursive-self-improvement Region:us Rsi Safetensors Self-evolving Self-improvement Self-improving-ai Sharded Supergpqa System-2 Tensorflow Thinking Typed-decisions Vidraft

Model Card on HF 🤗: https://huggingface.co/FINAL-Bench/Darwin-27B-RSI

Darwin 27B RSI Parameters and Internals

LLM Name	Darwin 27B RSI
Repository 🤗	https://huggingface.co/FINAL-Bench/Darwin-27B-RSI
Base Model(s)	FINAL-Bench/Darwin-27B-Opus FINAL-Bench/Darwin-27B-Opus
Model Size	27b
Required VRAM	53.8 GB
Updated	2026-09-29
Maintainer	FINAL-Bench
Model Type	qwen3_5_text
Model Files	49.8 GB: 1-of-2 4.0 GB: 2-of-2
Supported Languages	en ko
Model Architecture	Qwen3_5ForCausalLM
License	apache-2.0
Context Length	262144
Model Max Length	262144
Transformers Version	5.17.0
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	248320
Errors	replace

Rank the Darwin 27B RSI 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 Darwin 27B RSI

Best Alternatives	Context / RAM	Downloads	Likes
Hemmingway 1	256K / 53.8 GB	7880	773
OrcaSAQ 2 27B	256K / 12.3 GB	1330	169
Qwen3.6 27B OBLITERATED	256K / 53.6 GB	15137	184
Mnma Qwen3.8 27b Nvfp4	256K / 18.8 GB	527	11
Sharona Q27B R CodeSecurity V2	256K / 53.8 GB	1191	1
Qwen3.8 27B Code Tools Merged	256K / 53.8 GB	462	3
Qwen3.8 27B Obliterated E03	256K / 53.6 GB	708	1
...2 Abliterated OptiQ 3.7bpw Mlx	256K / 13.4 GB	6727	14
...wen3.8 P44w75 16.8B Unrepaired	256K / 2.4 GB	929	1
...2 Abliterated OptiQ 3.7bpw Mlx	256K / 13.4 GB	5919	14

Note: green Score (e.g. "73.2") means that the model is better than FINAL-Bench/Darwin-27B-RSI.