Model library / summerAIO
Qwen3.8-27B-Longbook27-v2 is a 27-billion-parameter model fine-tuned for long-form text understanding and generation.
Parameters in total
Context length in tokens, as published
Weights at the released precision
Licence declared on the repository
Overview
Qwen3.8-27B-Longbook27-v2 is a 27 billion-parameter language model released by summerAIO. The publisher describes it as a fine-tuned version of the Qwen 3.5 architecture, optimised for processing book-scale inputs such as long documents, papers and reports. It uses a custom Gated DeltaNet linear-attention design and is intended for summarisation, question answering and long-form generation tasks.
In practice the model occupies 27.1 billion parameters and requires loading the full model onto a GPU; CPU offloading is unsupported due to custom kernels. It supports a context window of 262 144 tokens, is distributed in bfloat16 weight format, and is released under the Apache-2.0 licence. Loading demands `trust_remote_code=True` because of repository-provided code.
Specification
| Repository | summerAIO/Qwen3.8-27B-Longbook27-v2 |
| Publisher | summerAIO |
| Published | 22 October 2026 |
| Architecture | Qwen35GDN24ForCausalLM |
| Model type | qwen3_5_gdn24 |
| Parameters | 27,072,285,696 |
| Layers | 64 |
| Hidden size | 5,120 |
| Attention heads | 24, 4 key/value heads |
| Context length | 262,144 tokens |
| Vocabulary | 248,320 tokens |
| Weight format | bfloat16 |
| Licence | apache-2.0 |
Memory
Parameter count multiplied by the bytes each format uses. The key-value cache comes on top.
| Precision | Weights | Notes |
|---|---|---|
| bf16 | 50.4 GiB | Full precision as released |
| fp8 | 25.2 GiB | 8-bit, near-lossless on most models |
| int4 | 12.6 GiB | 4-bit, smallest footprint |
The cache costs 256 KiB per token at 16-bit, so the full 262,144-token context of one request needs about 64.0 GiB. Concurrency multiplies that number, not the weights.
Hardware
At the released precision, with 90% of the card given to the server and 2.5 GiB kept for the runtime. The context column is what is left for the cache on that setup.
| GPU | Cards needed | Free for cache | Context it holds |
|---|---|---|---|
| RTX 4090 24GB | 4 | 26.0 GiB | about 106,389 tokens |
| RTX 5090 32GB | 2 | 2.2 GiB | about 8,904 tokens |
| L40S 48GB | 2 | 31.0 GiB | about 126,869 tokens |
| A100 80GB | 1 | 19.1 GiB | about 78,126 tokens |
| H100 80GB | 1 | 19.1 GiB | about 78,126 tokens |
| RTX PRO 6000 96GB | 1 | 33.5 GiB | about 137,109 tokens |
| H200 141GB | 1 | 74.0 GiB | the full 262,144 tokens |
Serving
A vLLM launch line for the shape above. Check the model card for a runtime the publisher recommends.
vllm serve summerAIO/Qwen3.8-27B-Longbook27-v2 \ --max-model-len 102400 \ --gpu-memory-utilization 0.90
Questions
The weights take about 50 GiB at the released precision. That fits on one A100 80GB, which leaves roughly 19 GiB for the key-value cache.
The repository declares apache-2.0. Read the licence text before commercial use: the name alone does not tell you what is allowed.
The configuration allows 262,144 tokens. Whether the whole window is usable depends on the memory left for the cache, which the hardware table works out per card.
This page is a reference, and Aiqre does not serve this model today. We deploy open-weight models on dedicated EU hardware on request, so if you want an OpenAI-compatible endpoint for it, email [email protected] with the model name and roughly what volume you expect.
Also in the library
Hemmingway-1.5 is a 26.9 B open-weight language model aimed at everyday writing tasks.
Nanbeige4.3-33B is a 33.1 billion-parameter Llama-style causal model released by Nanbeige for high-capacity agentic tasks and tea…
Nanbeige4.3-33B-Base is a 33.1-billion-parameter Llama-style causal language model released by Nanbeige under Apache-2.0.
Flint is a 39.8 billion-parameter language model for coding agents, using about 2.5 billion active parameters per token.
Hosting
Tell us the model and roughly what volume you expect. We reply with a price and an OpenAI-compatible endpoint, with your prompts never stored.