Roboflow
Qwen

Qwen: Qwen3.8 27B

Qwen3.8 27B Overview

Qwen3.8-27B is a dense vision-language model of roughly 27.78 billion parameters from Alibaba's Qwen team, built on the architectural foundation established by the Qwen3.5 generation. It takes interleaved text, image, and video input through an integrated vision encoder and produces text, covering document and chart understanding, visual question answering, front-end code generation from visual references, and agentic computer-use workflows. The decoder is a hybrid stack of 64 layers that repeats a pattern of three Gated DeltaNet linear-attention blocks, each followed by a feed-forward block, then one grouped-query full-attention block, so only a quarter of the layers maintain a conventional key-value cache. Native context length is 262,144 tokens.

Post-training targets long-horizon agentic work, and the model exposes per-request thinking control that lets callers switch reasoning traces on or off and choose a reasoning effort level. Multi-token prediction weights are trained alongside the main network for speculative decoding. Qwen reports improvements over Qwen3.6-27B on agentic and multimodal evaluations including Terminal-Bench 2.1, OSWorld-Verified, and SWE-bench Multimodal, with additional results on MathVision, CharXiv, and an internal Vision2Web suite.

Qwen3.8 27B Details & Performance

Details

Resources

Vision Tasks

CaptioningChart Question AnsweringClassificationDocument Question AnsweringImage TaggingMulti-Label ClassificationOCRVision LanguageVisual Question Answering

Features

Foundation VisionLLMs with Vision CapabilitiesMultimodal Vision

Usage

Past 30 Days

Not available

Not in Playground

Performance

Avg. Latency

Arena Rankings

Not yet ranked in arena

Qwen3.8 27B Vision Evals

Vision Evals is Roboflow's ground-truth benchmark: every model runs the same real-world samples across six vision tasks, and answers are scored against ground truth.

Evals updated August 14, 2026Pricing updated August 15, 2026

Overall score#29 of 30
61.2%
Avg cost / sample#9 of 30
$0.0018
Avg speed / sample#14 of 30
7.33s
Avg tokens / sample
2.2K

Strengths and weaknesses

Qwen3.8 27B averages 61.2% across the six Vision Evals tasks, ranking #29 of 30 models overall.

Its weakest relative showing is OCR, ranking #30 of 30 at 81.4%.

At $0.0018 per sample it is the 9th cheapest of the 30 benchmarked models, and its average inference time of 7.3s per sample makes it the 14th fastest.

Performance profile

Field medianQwen3.8 27B

Field medians: Object Detection 55.2%, Counting 64.2%, Identification 84.4%, OCR 90.0%, Data Extraction 86.6%, Reasoning 58.0%.

Results by task

TaskScoreField (0 to 100)RankCost / sampleSpeed
Object Detection
54.5%
#16 of 30$0.003615.64s
Counting
41.9%
#30 of 30$0.00051.07s
Identification
78.1%
#22 of 30$0.00051.13s
OCR
81.4%
#30 of 30$0.001910.57s
Data Extraction
79.4%
#26 of 30$0.00051.24s
Reasoning (low)
31.8%
#27 of 30$0.00051.08s
Reasoning (high)
62.3%
#19 of 30$0.008753.85s
  • Thinking longer helps: 30.5 points higher on reasoning at high effort for 17.8x the cost and 50x the latency.

Price vs. performance

Score vs. cost

Overall benchmark score against estimated cost per sample. Upper-left is the sweet spot: high quality at low cost.

30 models on the current benchmark · scores and efficiency pooled across all six tasks at low effort · Qwen3.8 27B highlighted

Qwen3.8 27B scores from a single evaluation run · Methodology

View all Vision Evals →

Qwen3.8 27B Pricing

Qwen3.8 27B costs $0.450 per 1M input tokens and $3.20 per 1M output tokens.

Input$0.450 / 1M tokens
Output$3.20 / 1M tokens

Pricing updated Aug 15, 2026

Qwen3.8 27B License

Apache-2.0 · Permissive license

Qwen3.8 27B is released under Apache-2.0, a permissive license. The Qwen3.8 27B license lets you run, fine-tune, and redistribute the model in commercial products with no obligation to open-source related code changes, so no separate commercial license is required.

Commercial use
Permitted. Because Apache-2.0 is permissive, Qwen3.8 27B can ship inside paid products and internal systems with no commercial license and no revenue threshold.
Modification
Permitted. Fine-tuning, quantizing, and distilling are all allowed, and your code changes can stay closed. Files you change must be marked as changed.
Redistribution
Permitted with attribution. Ship the Apache-2.0 license text and any NOTICE file alongside the weights or derived code.

Apache-2.0 grants an express patent license that terminates if you bring a patent claim over the work, and it disclaims warranties. Validate Qwen3.8 27B on your own data before you depend on it in production.

Read the full Apache 2.0 license ↗

Do I need a commercial license for Qwen3.8 27B?

This is the straightforward case: a permissive license is the best technical solution and you are free to deploy Qwen3.8 27B commercially without open-sourcing your own code.

Do not hesitate to reach out with questions for your commercial project — our team will help you start solving business problems on the first call. See Roboflow commercial licensing for the models included in each plan.

Talk to sales

This model is released under the Apache License 2.0, a permissive open-source license that allows commercial use, modification, distribution, and patent use.

Yes. Under the terms of the Apache 2.0 license, you can freely use this model for commercial purposes, including in proprietary products. You must retain the copyright notice and disclaimers when redistributing.

License information is provided as a guide and is not legal advice.

Frequently Asked Questions About Qwen3.8 27B Vision

Yes. Qwen3.8 27B accepts image input and handles OCR, data extraction, object counting, identification, visual reasoning, and object detection. On Roboflow's Vision Evals its strongest task is Object Detection at 54.5% (#16 of 30).

Yes. its transcriptions match the ground truth 81.4% on average (#30 of 30) on Vision Evals OCR. Pulling specific fields out of documents (data extraction) scores 79.4%.

Not its strength. On Vision Evals, Qwen3.8 27B scores 54.5% mAP@50 on object detection (#16 of 30) and 41.9% exact-match accuracy on object counting. For production counting or precise localization, pairing it with a specialized detector like RF-DETR or your own trained model in a Roboflow Workflow is usually more reliable: detect the objects, then count the detections.

On our benchmark's task mix, Qwen3.8 27B averages $0.0018 per sample at $0.45 per 1M input and $3.20 per 1M output tokens (#9 of 30 on cost), with an average speed of 7.3s per sample across the benchmark. Actual cost depends on your images and prompts.

On the overall Vision Evals ranking, Qwen3.8 27B sits #29 of 30 at 61.2%, just behind Qwen 3.7 Flash (61.7%) and just ahead of Kimi K2.6 (59%). See the full side-by-side: Qwen3.8 27B vs Qwen 3.7 Flash.