Qwen3.8-27B Benchmark Details
Qwen3.8-27B currently shows benchmark results led by LiveCodeBench (6 / 124, score 90.30), OSWorld-Verified (3 / 26, score 84.30), CharXiv RQ (2 / 15, score 90.20). This page also compares it with 4 competitor models and 4 predecessor or same-series models, including performance and pricing views when available.
Benchmark Results
Benchmark Results
Coding and Software Engineer
4 evaluationsAI Agent - Tool Usage
2 evaluationsMultimodal Understanding
7 evaluationsCompetitor Comparison
Benchmark scores for Qwen3.8-27B compared against top models in its class
10 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.
| Benchmark | Qwen3.8-27BCurrent | Claude Sonnet 5 | GPT-5.6 Terra | Gemini 3.7 Flash | DeepSeek-V4-Flash |
|---|---|---|---|---|---|
HLE 综合评估 | 30.80Thinking Enabled | 57.40Thinking Level · Extra High ｜ Tools | -- | -- | 45.10Thinking Level · Extra High ｜ Tools |
GPQA Diamond 科学与综合推理 | 89.20Thinking Enabled | 90.53Thinking Level · Extra High | 93.31Thinking Level · High | -- | 88.10Thinking Level · High |
DeepSWE 编程与软件工程 | 42.20Thinking Enabled ｜ Tools | 54.00Deep Thinking Mode ｜ Tools | 69.60Thinking Level · Extra High ｜ Tools | 65.30Thinking Level · High ｜ Tools | 54.40Thinking Level · High ｜ Tools |
LiveCodeBench 编程与软件工程 | 90.30Thinking Enabled | -- | -- | -- | 91.60Thinking Level · High |
NL2Repo-Bench 编程与软件工程 | 42.30Thinking Enabled ｜ Tools | -- | -- | -- | 54.20Thinking Level · High ｜ Tools |
SWE-Bench Pro - Public 编程与软件工程 | 61.70Thinking Enabled ｜ Tools | -- | -- | -- | 52.60Thinking Level · Extra High ｜ Tools |
OSWorld-Verified AI Agent - 工具使用 | 84.30Thinking Enabled ｜ Tools | 81.20Thinking Level · Extra High ｜ Tools | -- | -- | -- |
Terminal-Bench 2.1 AI Agent - 工具使用 | 73.00Thinking Enabled ｜ Tools | 80.40Thinking Level · Extra High ｜ Tools | 87.40Thinking Level · High | 85.80Thinking Enabled ｜ Tools | 82.70Thinking Level · High ｜ Tools |
Agents' Last Exam Agent能力评测 | 20.40Thinking Enabled ｜ Tools | -- | 50.40Thinking Level · Extra High ｜ Tools | 26.30Thinking Level · Medium ｜ Tools | 25.20Thinking Level · High ｜ Tools |
CharXiv RQ 多模态理解 | 90.20Thinking Enabled ｜ Tools | -- | -- | 88.70Thinking Level · Medium ｜ Tools | -- |
Standard API Pricing: Qwen3.8-27B vs. Peer Models
Shows standard text input and output pricing side by side for each model. If extended-context pricing exists, the chart keeps the base rate and explains the threshold below.
Source: DataLearnerAI. Standard text prices shown here use the default supplier. · USD / 1M tokens
| Model | Supplier | Standard input | Standard output | Base price applies to |
|---|---|---|---|---|
Claude Sonnet 5 | Anthropic | $2 / 1M tokens | $10 / 1M tokens | — |
GPT-5.6 Terra | OpenAI | $2.5 / 1M tokens | $15 / 1M tokens | — |
Gemini 3.7 Flash | DeepMind | $0.75 / 1M tokens | $3.75 / 1M tokens | — |
DeepSeek-V4-Flash | DeepSeek-AI | $0.14 / 1M tokens | $0.28 / 1M tokens | — |
Version History
How each version of the Qwen3.8-27B series stacks up on benchmark tests
5 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.· Click a row to view its trend chart.
| Benchmark | Qwen3.8-27BCurrent | Qwen3.6-27B | Qwen3.5-27B | Qwen3-32B | Qwen2.5-32B |
|---|---|---|---|---|---|
HLE 综合评估 | 30.80Thinking Enabled | 24.00Thinking Enabled | 48.50Thinking Enabled ｜ Tools | -- | -- |
GPQA Diamond 科学与综合推理 | 89.20Thinking Enabled | 87.80Thinking Enabled | 85.50Thinking Enabled | 68.40Thinking Enabled | -- |
LiveCodeBench 编程与软件工程 | 90.30Thinking Enabled | 83.90Thinking Enabled | 80.70Thinking Enabled ｜ Tools | 65.70Thinking Enabled | 51.20Standard Mode |
SWE-Bench Pro - Public 编程与软件工程 | 61.70Thinking Enabled ｜ Tools | 53.50Thinking Enabled ｜ Tools | -- | -- | -- |
OSWorld-Verified AI Agent - 工具使用 | 84.30Thinking Enabled ｜ Tools | -- | 56.20Thinking Enabled ｜ Tools | -- | -- |
Single-Benchmark Version Trend
Viewing: HLE · 综合评估
Standard API Pricing Across the Qwen3.8-27B Series
Shows standard text input and output pricing side by side for each model. If extended-context pricing exists, the chart keeps the base rate and explains the threshold below.
Source: DataLearnerAI. Standard text prices shown here use the default supplier.
These models use different currencies or billing units, so the page falls back to raw price values instead of a shared bar chart.
| Model | Supplier | Standard input | Standard output | Base price applies to |
|---|---|---|---|---|
Qwen3-32B | 阿里巴巴 | ¥0.0012 / 1K tokens | ¥0.0048 / 1K tokens | — |
Qwen2.5-32B | 阿里巴巴 | ¥0.002 / 1K tokens | ¥0.006 / 1K tokens | — |