Alibaba Cloud · Apache 2.0 / Qwen
Qwen3.8 API
Alibaba's most capable family — multimodal flagship plus an open dense model.
Context: up to 1M tokens · Released: August 2026
# Qwen3.8 via kryyt (OpenAI SDK) from openai import OpenAI client = OpenAI(base_url="https://api.kryyt.com/v1", api_key="$KRYYT_API_KEY") r = client.chat.completions.create( model="qwen3.8-max", messages=[{"role":"user","content":"Explain MoE routing in one paragraph."}], ) print(r.choices[0].message.content)
Overview
Qwen3.8 is Alibaba's newest generation. Qwen3.8-Max is a 2.4-trillion-parameter mixture-of-experts flagship with a 1M-token context and native vision; Qwen3.8-27B is a compact, Apache-2.0 dense model with vision-language input, ideal for self-hosting and cost-efficient serving.
Best for:
- Multimodal understanding (text, image, video)
- Multilingual coverage
- Cost-efficient dense deployment
Qwen3.8 benchmark performance
Qwen3.8-Max stands out on research reproduction (PaperBench), computer use (OSWorld-Verified), and instruction following, while trailing on SWE-bench Pro.
| Benchmark | Model | Score (%) |
|---|---|---|
| PaperBenchReproducing ML research papers | Qwen3.8-Max | |
| GPQA DiamondGraduate-level science reasoning | Qwen3.8-Max | |
| Terminal-Bench 2.1Agentic coding in real terminal environments | Qwen3.8-Max | |
| OSWorld-VerifiedOperating a real desktop computer | Qwen3.8-Max | |
| IFBenchPrecise instruction following | Qwen3.8-Max | |
| SWE-bench ProResolving issues in professional codebases | Qwen3.8-Max | |
| Humanity's Last ExamFrontier knowledge and reasoning (no tools) | Qwen3.8-Max |
Sources: Qwen 3.8 benchmarks (Apidog, citing Alibaba) (opens in new tab); Qwen3.8-Max benchmarks (DataCamp, citing Alibaba) (opens in new tab).
Qwen3.8-Max vs Qwen3.7-Max
| Benchmark | Scores (%) |
|---|---|
| SWE-bench ProResolving issues in professional codebases | |
| IFBenchPrecise instruction following | |
| Humanity's Last ExamFrontier knowledge and reasoning (no tools) |
Sources: Qwen 3.8 benchmarks (Apidog, citing Alibaba) (opens in new tab).
Scores are vendor-reported unless marked independent, as of September 2026. Labs use different agent harnesses, reasoning settings, and benchmark versions, so treat cross-lab gaps of a few points as roughly equal and test on your own workload. See all models on the benchmarks page.
Qwen3.8 models & pricing
| Model | API ID | Context | Pricing | License | Status |
|---|---|---|---|---|---|
| Qwen3.8-Max | qwen3.8-max | 1M | See account pricing | Apache 2.0 / Qwen | Soon |
| Qwen3.8-27B | qwen3.8-27b | 262K | See account pricing | Apache 2.0 / Qwen | Soon |
| Qwen3.8-Flash | qwen3.8-flash | 256K | See account pricing | Apache 2.0 / Qwen | Soon |
qwen3.8-max
- Model
- Qwen3.8-Max
- Context
- 1M
- Pricing
- See account pricing
- License
- Apache 2.0 / Qwen
- Status
- Soon
qwen3.8-27b
- Model
- Qwen3.8-27B
- Context
- 262K
- Pricing
- See account pricing
- License
- Apache 2.0 / Qwen
- Status
- Soon
qwen3.8-flash
- Model
- Qwen3.8-Flash
- Context
- 256K
- Pricing
- See account pricing
- License
- Apache 2.0 / Qwen
- Status
- Soon
Usage is charged at Kryyt's published rates for the selected model and features. Review current pricing and usage in your account.
Qwen3.8 FAQ
What is Qwen3.8?
Qwen3.8 is Alibaba's newest generation. Qwen3.8-Max is a 2.4-trillion-parameter mixture-of-experts flagship with a 1M-token context and native vision; Qwen3.8-27B is a compact, Apache-2.0 dense model with vision-language input, ideal for self-hosting and cost-efficient serving.
How do I call Qwen3.8 on kryyt?
Send a POST request to https://api.kryyt.com/v1/chat/completions with your kryyt key and set "model" to one of: qwen3.8-max, qwen3.8-27b, qwen3.8-flash.
What license is Qwen3.8 under?
Qwen3.8 is available under the Apache 2.0 / Qwen license via its provider, Alibaba Cloud.
What context window does Qwen3.8 support?
Qwen3.8 supports up to 1M tokens depending on the specific model variant.
Ship with one endpoint
Get a key, change your base URL, and call every model on this page. No new SDKs.