Zyphra/ZAYA1-8B · Hugging Face

Source: https://huggingface.co/Zyphra/ZAYA1-8B ZAYA1-8B is a small mixture of experts language model with 760M active parameters and 8.4B total parameters trained end-to-end by Zyphra. ZAYA1-8B sets a new standard of intelligence efficiency for its parameter count through a combination of novel architecture and innovations in pretraining and post-training.

ZAYA1-8B excels at detailed long-form reasoning especially for mathematical and coding task. It punches heavily above its weight in these regimes and due to its inference efficiency and small size can be highly effective in test-time compute harnesses.

Due to its small total parameter count, ZAYA1-8B can also be deployed on-device for local LLM applications.

Learn more in ourtechnical reportandblog.

This is the post-trained reasoning version of ZAYA1-8B. The pretraining base can be foundhere.

https://huggingface.co/Zyphra/ZAYA1-8B#performancePerformance

ZAYA1-8B performs extremely strongly, especially in challenging mathematical, reasoning, and coding benchmarks. ZAYA1-8B is competitive with models several times its own size including frontier-scale reasoning models at mathematical benchmarks.

First we compare ZAYA1-8B to the SOTA Qwen3 and Qwen3.5 model series of approximately the same parameter count as well as the recently released Gemma4 models and secondly to a variety of larger open-weights models.

https://huggingface.co/Zyphra/ZAYA1-8B#in-class-comparison-against-open-source-reasoning-modelsIn-class comparison against open-source reasoning models

CategoryBenchmarkZAYA1-8B (0.7B / 8.0B)Qwen3-4B-Thinking-2507 (4.0B / 4.0B)Qwen3.5-4B (4.0B / 4.0B)Gemma-4-E4B-it (4.0B / 8.0B*)MathAIME’2689.177.584.550.3MathHMMT Feb.’2671.660.863.632.1MathIMO-AnswerBench59.350.948.727.3MathAPEX-shortlist32.216.9--6.1CodeLiveCodeBench-v665.854.2--54.2KnowledgeGPQA-Diamond71.066.576.257.4KnowledgeMMLU-Pro74.274.379.170.2InstructionIFEval85.5886.889.888.50InstructionIFBench52.5652.959.242.67Style & chatEQBench72.9579.679.580.15Style & chatCreative Writing v362.9758.672.983.75AgenticBFCL-v439.2249.745.231.7Agenticτ²43.1252.982.137.7

https://huggingface.co/Zyphra/ZAYA1-8B#scaling-comparison-against-larger-open-source-reasoning-modelsScaling comparison against larger open-source reasoning models

ModelActiveTotalAIME’26HMMT’26LCB-v6IFEvalGPQA-DMMLU-ProZAYA1-8B0.7B8B89.171.663.885.871.074.2Arcee-Trinity-Mini3B26B59.636.933.362.046.870.6N3-Nano-30B3B30B90.175.564.692.875.178.9OLMo-3.1-32B-Think32B32B78.950.658.393.259.675.8Qwen3-Next-80B-A3B-Think3B80B90.279.367.888.576.782.6Intellect-312B106B86.372.266.881.274.682.3Mistral-Small-4-119B6B119B86.470.657.984.077.281.6 All numbers are run on the Zyphra evaluation harness. Models are ordered by total parameter count.

https://huggingface.co/Zyphra/ZAYA1-8B#quickstartQuickstart

https://huggingface.co/Zyphra/ZAYA1-8B#prerequisitesPrerequisites

We recommend installing the following libraries in a fresh python environment (tested with python 3.12).

To use ZAYA1-8B, installzaya1\-prbranch from our fork ofvllmlibrary (the command will trigger a full build of vLLM from source):

pip install "vllm @ git+https://github.com/Zyphra/vllm.git@zaya1-pr"

If you want to run in transformers, installzaya1branch from our fork oftransformerslibrary as well:

pip install "transformers @ git+https://github.com/Zyphra/transformers.git@zaya1"

https://huggingface.co/Zyphra/ZAYA1-8B#deploymentDeployment

To start vLLM server, run the following command:

vllm serve Zyphra/ZAYA1-8B --port 8010 \
   --mamba-cache-dtype float32 --dtype bfloat16 \
   --reasoning-parser qwen3 --enable-auto-tool-choice --tool-call-parser zaya_xml

For parallel deployment we recommend using DP with EP as TP for CCA is not supported in the branch above. If running on 8 GPUs, set extra flags\-dp 8 \-epto run with DP=EP=8.

For our evaluations and for general use, we recommend temperature 1.0, top-p 0.95, top-k -1. For agent and code use cases, we recommend top-p 0.6.

Once the server is up, you can query a model withcurllike in the following example:

curl http://localhost:8010/v1/chat/completions \
    -H "Content-Type: application/json" \
    -d '{
        "model": "Zyphra/ZAYA1-8B",
        "messages": [
            {"role": "system", "content": "You are a helpful assistant."},
            {"role": "user", "content": "Hello. How is it going?"}
        ]
    }'