Pyyan / Compare / Qwen3.8-Omni-Flash vs GPT-6 Astra
Qwen3.8-Omni-Flash vs GPT-6 Astra
2 of 5Language Models · verified 20 Sept 2026
Qwen3.8-Omni-FlashAlibabacurrent
GPT-6 AstraOpenAIcurrent| Specification | Qwen3.8-Omni-Flash | GPT-6 Astra |
|---|---|---|
| Summary | Text, image, audio and video into one million tokens of context. | Saturates three of the benchmarks it reports, and ships disabled for enterprise. |
| Context | 1M | 1.05M |
| Input ($/Mtok) | $0.15 | $10 |
| Output ($/Mtok) | $0.47 | $50 |
| Licence | Proprietary | Proprietary |
| Max output | Not published | 128K |
| Parameters | Not published | Not published |
| Released | 18 Sep 2026 | 3 Sep 2026 |
| Model ID | qwen3.8-omni-flash | gpt-6-astra |
| Category | Language Models | Language Models |
| Official | Alibaba ↗ | — |
Highlighted rows are where these differ.
Qwen3.8-Omni-Flash
- Takes text, images, audio and video in one pass and returns text
- Alibaba reports more than 26% average gain over Qwen3.5-Omni-Plus across 30 evaluations
- No weights published, which is a change of habit for this team
Best for audio and video agents.
GPT-6 Astra
- 97.6% on FrontierMath Tier 4, 99.9% on ARC-AGI-3 and 100% on ExploitBench
- Above 272K input tokens the whole request reprices: input and cache at 2x, output at 1.5x
- Rolls out to OpenAI's Daybreak cyber programme first, and arrives switched off for enterprise
Best for hard reasoning and computer use.