InferX Catalog | Qwen3.5-35B-A3B-FP8

Qwen3.5-35B-A3B-FP8

Optimized with FP8 for efficient, high-performance reasoning and chat workloads

Qwen text text2text high-throughput reasoning coding agentic

Log in to deploy

Qwen/Qwen3.5-35B-A3B-FP8 is a 35B parameter Qwen3.5 model using A3B architecture and FP8 quantization to reduce memory usage and increase throughput while maintaining strong reasoning, coding, and agent capabilities, making it well-suited for high-performance production deployments.

Metadata

Provider

Qwen

Modality

text

API type

text2text

Source

huggingface / Qwen/Qwen3.5-35B-A3B-FP8

Created

2026-03-31 16:49:39 UTC

Updated

2026-05-11 21:46:05 UTC

Catalog version

3

Visibility

Published

Specifications

Parameters

35.00B

MoE

No

Max model length

262144

Image

vllm/vllm-openai:v0.20.2-cu129

Default Deploy Config

GPU count

1

vRAM

70000 MB

Summary

1xGPU 70000 MB

Recommended Use Cases

Reasoning assistant
Coding assistant
Agent / tool use

Model Spec

{
    "image": "vllm/vllm-openai:v0.20.2-cu129",
    "commands": [
        "--model",
        "Qwen/Qwen3.5-35B-A3B-FP8",
        "--trust-remote-code",
        "--gpu-memory-utilization",
        "0.92",
        "--max-model-len",
        "262144",
        "--reasoning-parser",
        "qwen3",
        "--enable-auto-tool-choice",
        "--tool-call-parser",
        "qwen3_coder",
        "--speculative-config",
        "{\"model\":\"Qwen/Qwen3.6-35B-A3B-FP8\",\"num_speculative_tokens\":4}",
        "--tensor-parallel-size=1"
    ],
    "resources": {
        "GPU": {
            "Count": 1,
            "vRam": 70000
        }
    },
    "envs": [
        [
            "VLLM_USE_DEEP_GEMM",
            "0"
        ]
    ],
    "policy": {
        "Obj": {
            "min_replica": 0,
            "max_replica": 1,
            "standby_per_node": 1,
            "parallel": 50,
            "queue_len": 100,
            "queue_timeout": 30.0,
            "scalein_timeout": 1.0,
            "scaleout_policy": {
                "WaitQueueRatio": {
                    "wait_ratio": 0.1
                }
            },
            "runtime_config": {
                "graph_sync": false
            }
        }
    },
    "sample_query": {
        "body": {
            "model": "Qwen/Qwen3.5-35B-A3B-FP8",
            "stream": "true",
            "max_tokens": "1000",
            "temperature": "0"
        },
        "path": "v1/completions",
        "prompt": "write a quick sort algorithm.",
        "apiType": "text2text",
        "dataUrl": "",
        "prompts": [],
        "loadingTimeout": 90
    }
}