Qwen3.5 397B-A17B
- Context
- 256K
- 66K max output
- Input price
- $0.17
- per 1M tokens · Alibaba Cloud Model Studio
- Output price
- $1.03
- per 1M tokens
- Epoch Capabilities Index
- 146.6
- #54 of 71 listed
- LMArena Text
- 1,442
- #48 of 79 listed
Where to run it
3 offerings · prices as listed| Meter | Tier | Price |
|---|---|---|
| Input | Standard | $0.17 / 1M tokens |
| Prompts over 128K tokens | $0.43 / 1M tokens | |
| Output | Standard | $1.03 / 1M tokens |
| Prompts over 128K tokens | $2.58 / 1M tokens | |
| Input | Standard | $0.60 / 1M tokens |
| Output | Standard | $3.60 / 1M tokens |
| Input | Standard | ¥1.20 / 1M tokens |
| Prompts over 128K tokens | ¥3.00 / 1M tokens | |
| Output | Standard | ¥7.20 / 1M tokens |
| Prompts over 128K tokens | ¥18.00 / 1M tokens | |
Changes
All changes →- 1 price dropped on Alibaba Cloud Model Studio Output (Thinking on) $1.03
- 2 prices dropped on Alibaba Cloud Model Studio, International Output (Thinking off) $3.60 · Output (Thinking on) $3.60
Vendor-reported scores
Published by Alibaba Qwen, under the conditions it states| Benchmark | Setting | Score |
|---|---|---|
| MMLU-Pro | thinking mode | 87.8% |
| GPQA Diamond | thinking mode | 88.4% |
| Humanity's Last Exam | no tools, thinking mode | 28.7% |
| with tools (search agent, context folding) | 48.3% | |
| LiveCodeBench | v6, thinking mode | 83.6% |
| τ²-bench | airline domain with Claude Opus 4.5 system-card fixes | 86.7% |
| BrowseComp | simple context folding | 69% |
| discard-all context strategy | 78.6% | |
| SWE-bench Verified | thinking mode | 76.4% |
| Terminal-Bench 2.0 | thinking mode | 52.5% |
| MMMU | thinking mode | 85% |
Cited scores
From third-party leaderboards, not measured by us- Epoch Capabilities Index 146.6
#54 of 71 listed models · CI 144.8–148.2
Epoch AI, 'Capabilities & benchmarking'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/benchmarks'. Licensed CC BY 4.0. · License · as of Sep 26, 2026 · changes we made
- LMArena Text 1,442
#48 of 79 listed models · CI 1,439–1,445 · 86,319 votes
LMArena (Arena), leaderboard-dataset on Hugging Face, licensed CC BY 4.0 · License · as of Sep 25, 2026 · changes we made
- Epoch AI · FrontierMath Tiers 1–3 31.2%
#48 of 56 listed models · reasoning off · ±2.75
Also listed: qwen3.5-397b-a17b 29.5%
Epoch AI, 'Capabilities & benchmarking'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/benchmarks'. Licensed CC BY 4.0. · License · as of Sep 26, 2026 · changes we made
- Epoch AI · GPQA Diamond 86.4%
#42 of 64 listed models · reasoning off · ±2.45
Also listed: qwen3.5-397b-a17b 85.9%
Epoch AI, 'Capabilities & benchmarking'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/benchmarks'. Licensed CC BY 4.0. · License · as of Sep 26, 2026 · changes we made
- LMArena Text · Chinese 1,499
#36 of 79 listed models · CI 1,490–1,507 · 5,974 votes
LMArena (Arena), leaderboard-dataset on Hugging Face, licensed CC BY 4.0 · License · as of Sep 25, 2026 · changes we made
- LMArena Text · Coding 1,491
#48 of 79 listed models · CI 1,486–1,496 · 24,819 votes
LMArena (Arena), leaderboard-dataset on Hugging Face, licensed CC BY 4.0 · License · as of Sep 25, 2026 · changes we made
- LMArena Vision 1,265
#28 of 49 listed models · CI 1,259–1,271 · 27,569 votes
LMArena (Arena), leaderboard-dataset on Hugging Face, licensed CC BY 4.0 · License · as of Sep 13, 2026 · changes we made