Model library / InternScience
Sage-A1.5 is a 35-billion-parameter mixture-of-experts model aimed at long-horizon scientific and agentic tasks.
Mixture of experts, count not published
Context length in tokens, as published
Weights at the released precision
Licence declared on the repository
Overview
Sage-A1.5 is a Mixture-of-Experts model built on Qwen3.6-35B-A3B and Agents-A1, released by InternScience. The publisher describes it as an agentic system for scientific problem solving, offering capabilities such as agentic orchestration, self-verification, and open scientific discovery, with native tool-use and instruction following.
The model contains roughly 36 billion parameters and supports a context window of 262 144 tokens. It is released under the Apache-2.0 licence, but the weight format and expert count are not published. Compatibility with Hugging Face Transformers, vLLM and SGLang is provided, facilitating integration into existing pipelines.
Specification
| Repository | InternScience/Sage-A1.5 |
| Publisher | InternScience |
| Published | 15 October 2026 |
| Architecture | Qwen3_5MoeForConditionalGeneration |
| Model type | qwen3_5_moe |
| Parameters | 35,951,822,704 |
| Layers | 40 |
| Hidden size | 2,048 |
| Attention heads | 16, 2 key/value heads |
| Context length | 262,144 tokens |
| Vocabulary | 248,320 tokens |
| 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 | 67.0 GiB | Full precision as released |
| fp8 | 33.5 GiB | 8-bit, near-lossless on most models |
| int4 | 16.7 GiB | 4-bit, smallest footprint |
The cache costs 80 KiB per token at 16-bit, so the full 262,144-token context of one request needs about 20.0 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 | 9.4 GiB | about 123,660 tokens |
| RTX 5090 32GB | 4 | 38.2 GiB | the full 262,144 tokens |
| L40S 48GB | 2 | 14.4 GiB | about 189,196 tokens |
| A100 80GB | 1 | 2.5 GiB | about 33,220 tokens |
| H100 80GB | 1 | 2.5 GiB | about 33,220 tokens |
| RTX PRO 6000 96GB | 1 | 16.9 GiB | about 221,964 tokens |
| H200 141GB | 1 | 57.4 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 InternScience/Sage-A1.5 \ --max-model-len 122880 \ --gpu-memory-utilization 0.90
Questions
The weights take about 67 GiB at the released precision. That fits on one A100 80GB, which leaves roughly 3 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 hello@aiqre.com with the model name and roughly what volume you expect.
Also in the library
MLExpert is a 35 billion-parameter mixture-of-experts model for autonomous machine-learning engineering tasks.
Flint is a 39.8 billion-parameter language model for coding agents, using about 2.5 billion active parameters per token.
Hy-Memory-30B-A3B is a Tencent-developed long-term-memory writer model designed to replace LLMs in Mem0, MemOS and Hy-Memory fram…
Xing4.0-29B-Intent is a schema-guided large language model for intent recognition and structured information extraction.
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.