Model library / amazon
ALoDLM-8B is an 8.2 billion-parameter diffusion language model from Amazon for research on adaptive computation, maths, and code generation.
Parameters in total
Context length in tokens, as published
Weights at the released precision
Licence declared on the repository
Overview
ALoDLM-8B is a diffusion language model developed by Amazon. It generates text and code by repeatedly applying shared transformer layers, refining latent token states and adaptively allocating computation. The model is intended for non-commercial research on diffusion language modelling, adaptive computation, mathematical reasoning and code generation, and it supports inference-time control of computation via gate and entropy thresholds.
The model contains 8.2 billion parameters, uses the Qwen3ForCausalLM architecture without expert layers, and accepts up to 40,960 tokens of context. Weights are stored in bfloat16 format and released under a CC-BY-NC-4.0 licence. Deployment requires a CUDA-capable NVIDIA GPU that supports BF16, and inference can be run with either the optimized nano-vLLM engine or a portable PyTorch decoder.
Specification
| Repository | amazon/ALoDLM-8B |
| Publisher | amazon |
| Published | 2 October 2026 |
| Architecture | Qwen3ForCausalLM |
| Model type | qwen3 |
| Parameters | 8,190,735,360 |
| Layers | 36 |
| Hidden size | 4,096 |
| Attention heads | 32, 8 key/value heads |
| Context length | 40,960 tokens |
| Vocabulary | 151,936 tokens |
| Weight format | bfloat16 |
| Licence | cc-by-nc-4.0 |
Memory
Parameter count multiplied by the bytes each format uses. The key-value cache comes on top.
| Precision | Weights | Notes |
|---|---|---|
| bf16 | 15.3 GiB | Full precision as released |
| fp8 | 7.6 GiB | 8-bit, near-lossless on most models |
| int4 | 3.8 GiB | 4-bit, smallest footprint |
The cache costs 144 KiB per token at 16-bit, so the full 40,960-token context of one request needs about 5.6 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 | 1 | 3.8 GiB | about 27,987 tokens |
| RTX 5090 32GB | 1 | 11.0 GiB | the full 40,960 tokens |
| L40S 48GB | 1 | 25.4 GiB | the full 40,960 tokens |
| A100 80GB | 1 | 54.2 GiB | the full 40,960 tokens |
| H100 80GB | 1 | 54.2 GiB | the full 40,960 tokens |
| RTX PRO 6000 96GB | 1 | 68.6 GiB | the full 40,960 tokens |
| H200 141GB | 1 | 109.1 GiB | the full 40,960 tokens |
Serving
A vLLM launch line for the shape above. Check the model card for a runtime the publisher recommends.
vllm serve amazon/ALoDLM-8B \ --max-model-len 24576 \ --gpu-memory-utilization 0.90
Questions
The weights take about 15 GiB at the released precision. That fits on one RTX 4090 24GB, which leaves roughly 4 GiB for the key-value cache.
The repository declares cc-by-nc-4.0. Read the licence text before commercial use: the name alone does not tell you what is allowed.
The configuration allows 40,960 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
Granite Switch 4.1 8B is a modular 8.8 billion-parameter language model that combines a base LLM, decision and safety adapters, a…
EdgeRock-4B is a 4.6 billion-parameter multimodal language model with constant-memory reasoning capabilities.
EdgeRock-4B is a 4.6 billion-parameter multimodal language model with constant-size memory for text and image reasoning.
Mellum2.1 Thinking is a 12 billion-parameter mixture-of-experts language model designed for complex agentic coding, math and reas…
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.