Qwen

Qwen3.5-122B-A10B

Tool callingStructured outputReasoningImage inputAudio inputVideo input
Qwen3.5-122B-A10B model card cover

Qwen vision-language model for visual reasoning, documents, and agent tasks

Context
262K
Released
7mo ago
Parameters
122B
Input / 1M tokens
$0.26
Output / 1M tokens
$2.08

Sources:models.devOpenRouter

Benchmarks

DeepSWE is the primary coding benchmark where an exact model evaluation exists. Other sourced evaluations appear below; missing scores are not zero.

Not evaluated in DeepSWE 1.1 as of 2026-09-22. Results are only shown for exact model versions in the official dataset. Official leaderboard.

No benchmark results have been matched to this model yet.

Specifications

Developer
Qwen
Model family
Qwen
Type
Language model
Predecessor
—
Successor
—
Released
February 23, 2026
Model ID
qwen/qwen3.5-122b-a10b
Openness
Open weights
License
—
Parameters
122B
Input modalities
text, image, video
Output modalities
text
Modality
text+image+video->text
Native context
256K
Tokenizer
Qwen3
Max output tokens
66K
Source freshness
Metadata synced Oct 8, 2026, 10:17 PM UTC

Open-weight builds

Ollama variants

Every published Ollama tag matched to this model, including parameter editions, instruction/base variants, and quantized builds.

63 builds
VariantParametersTypeQuantizationSizeContextInput
qwen3.5:0.8b-bf160.8B—BF161.8 GB256Ktext + image
qwen3.5:0.8b-mlx-bf160.8B—BF161.7 GB256Ktext + image
qwen3.5:0.8b-mtp-q8_00.8B—Q8_01.3 GB256Ktext + image
qwen3.5:0.8b-q8_00.8B—Q8_01 GB256Ktext + image
qwen3.5:0.8b0.8B——1.2 GB256Ktext + image
qwen3.5:0.8b-mlx0.8B——1.2 GB256Ktext + image
qwen3.5:0.8b-nvfp40.8B——1 GB256Ktext + image
qwen3.5:2b-bf162B—BF164.6 GB256Ktext + image
qwen3.5:2b-mlx-bf162B—BF164.4 GB256Ktext + image
qwen3.5:2b-mtp-q4_K_M2B—Q4_K_M2.6 GB256Ktext + image
qwen3.5:2b-q4_K_M2B—Q4_K_M1.9 GB256Ktext + image
qwen3.5:2b-mtp-q8_02B—Q8_03.3 GB256Ktext + image
qwen3.5:2b-q8_02B—Q8_02.7 GB256Ktext + image
qwen3.5:2b2B——2.7 GB256Ktext + image
qwen3.5:2b-mlx2B——3.1 GB256Ktext + image
qwen3.5:2b-nvfp42B——2.5 GB256Ktext + image
qwen3.5:4b-bf164B—BF169.3 GB256Ktext + image
qwen3.5:4b-mlx-bf164B—BF169.1 GB256Ktext + image
qwen3.5:4b-mtp-q4_K_M4B—Q4_K_M4.1 GB256Ktext + image
qwen3.5:4b-q4_K_M4B—Q4_K_M3.3 GB256Ktext + image
qwen3.5:4b-mtp-q8_04B—Q8_06 GB256Ktext + image
qwen3.5:4b-q8_04B—Q8_05.2 GB256Ktext + image
qwen3.5:4b4B——3.3 GB256Ktext + image
qwen3.5:4b-mlx4B——4 GB256Ktext + image
qwen3.5:4b-mxfp84B——5.6 GB256Ktext + image
qwen3.5:9b-bf169B—BF1619 GB256Ktext + image
qwen3.5:9b-mlx-bf169B—BF1619 GB256Ktext + image
qwen3.5:9b-mtp-q4_K_M9B—Q4_K_M9 GB256Ktext + image
qwen3.5:9b-q4_K_M9B—Q4_K_M6.6 GB256Ktext + image
qwen3.5:9b-mtp-q8_09B—Q8_013 GB256Ktext + image
qwen3.5:9b-q8_09B—Q8_010 GB256Ktext + image
qwen3.5:9b-mlx9B——8.9 GB256Ktext + image
qwen3.5:9b-mtp-nvfp49B——7.6 GB256Ktext + image
qwen3.5:9b-mxfp89B——11 GB256Ktext + image
qwen3.5:9b-nvfp49B——7.2 GB256Ktext + image
qwen3.5:latest9B——6.6 GB256Ktext + image
qwen3.5:27b-bf1627B—BF1656 GB256Ktext + image
qwen3.5:27b-coding-bf1627B—BF1655 GB256Ktext + image
qwen3.5:27b-mlx-bf1627B—BF1655 GB256Ktext + image
qwen3.5:27b-mtp-q4_K_M27B—Q4_K_M21 GB256Ktext + image
qwen3.5:27b-q4_K_M27B—Q4_K_M17 GB256Ktext + image
qwen3.5:27b-q8_027B—Q8_030 GB256Ktext + image
qwen3.5:27b27B——17 GB256Ktext + image
qwen3.5:27b-coding-mxfp827B——30 GB256Ktext + image
qwen3.5:27b-coding-nvfp427B——20 GB256Ktext + image
qwen3.5:27b-int427B——16 GB256Ktext + image
qwen3.5:27b-int827B——30 GB256Ktext + image
qwen3.5:27b-mlx27B——20 GB256Ktext + image
qwen3.5:27b-mxfp827B——30 GB256Ktext + image
qwen3.5:35b-a3b-bf1635B—BF1672 GB256Ktext + image
qwen3.5:35b-a3b-coding-bf1635B—BF1670 GB256Ktext + image
qwen3.5:35b-a3b-mlx-bf1635B—BF1670 GB256Ktext + image
qwen3.5:35b-a3b-mtp-q4_K_M35B—Q4_K_M24 GB256Ktext + image
qwen3.5:35b-a3b-q4_K_M35B—Q4_K_M22 GB256Ktext + image
qwen3.5:35b-a3b-q8_035B—Q8_039 GB256Ktext + image
qwen3.5:35b35B——22 GB256Ktext + image
qwen3.5:35b-a3b-coding-mxfp835B——37 GB256Ktext + image
qwen3.5:35b-a3b-coding-nvfp435B——22 GB256Ktext + image
qwen3.5:35b-a3b-int435B——20 GB256Ktext + image
qwen3.5:35b-a3b-int835B——38 GB256Ktext + image
qwen3.5:35b-a3b-mxfp835B——37 GB256Ktext + image
qwen3.5:35b-a3b-nvfp435B——22 GB256Ktext + image
qwen3.5:122b122B—Q4_K_M81 GB256Ktext + image
Additional source evidence
Source model
alibaba/qwen3.5-122b-a10b
Source family
qwen
Last updated
2026-02-23
Model context
262K
Model output limit
66K
Open weights
Yes
File attachments
Yes
Reasoning
Yes
Tool calling
Yes
Structured output
Yes
Temperature control
Yes

Downloadable weights

Published benchmarks

BenchmarkScoreEvaluation detailsEvidence
SWE-Bench Verified72 · resolved—Source
Complete source record (JSON)
{
  "id": "alibaba/qwen3.5-122b-a10b",
  "name": "Qwen3.5 122B-A10B",
  "description": "Qwen vision-language model for visual reasoning, documents, and agent tasks",
  "family": "qwen",
  "last_updated": "2026-02-23",
  "attachment": true,
  "reasoning": true,
  "tool_call": true,
  "structured_output": true,
  "temperature": true,
  "modalities": {
    "input": [
      "text",
      "image",
      "video",
      "audio"
    ],
    "output": [
      "text"
    ]
  },
  "limit": {
    "context": 262144,
    "output": 65536
  },
  "benchmarks": [
    {
      "name": "SWE-Bench Verified",
      "score": 72,
      "source": "https://huggingface.co/Qwen/Qwen3.5-122B-A10B",
      "metric": "resolved"
    }
  ],
  "release_date": "2026-02-23",
  "open_weights": true,
  "weights": [
    {
      "label": "Hugging Face",
      "url": "https://huggingface.co/Qwen/Qwen3.5-122B-A10B"
    }
  ]
}

Infrastructure

Provider availability

All providers
ProviderEndpointContextInput /M tokensOutput /M tokensCached input /M tokensCache write /M tokensUptime 1dData handling
Alibabaalibaba262K$0.26$2.08——95.12%
ZDR: No

Training: Unknown

Prompt logging: Unknown

Retention: Unknown

Checked 2026-10-15

SiliconFlowsiliconflow/fp8262K$0.26$2.08——50.99%
ZDR: Yes

Training: Unknown

Prompt logging: Unknown

Retention: Unknown

Checked 2026-10-15

AtlasCloudatlas-cloud/fp8262K$0.30$2.40$0.30—100.00%
ZDR: No

Training: Unknown

Prompt logging: Unknown

Retention: Unknown

Checked 2026-10-15

Novitanovita/bf16262K$0.40$3.20——100.00%
ZDR: Yes

Training: Unknown

Prompt logging: Unknown

Retention: Unknown

Checked 2026-10-15

Provider data:OpenRouter

Deployment graph

Can I run this?

Vertical Tabs only shows compatibility when it has structured evidence tying this exact model to hardware, runtime, quantization, memory use, and observed performance.

Browse hardware