# DeepSeek V4.1 Flash vs Hy3 preview - AI Model Comparison

- Page: https://opencode.ai/data/compare/deepseek/deepseek-v4-1-flash/tencent/hy3-preview
- Updated: 2026-10-02T09:36:21.000Z

|  | DeepSeek V4.1 Flash | Hy3 preview |
| --- | --- | --- |
| Lab | DeepSeek | Tencent |
| Context window | 1M | 256K |
| Max output | 384K | 64K |
| Input modalities | text, image | text |
| Reasoning | Yes | Yes |
| Tool calling | Yes | Yes |
| Open weights | Yes | Yes |
| Release date | 2026-09-10 | 2026-04-20 |
| Input price per 1M tokens | $0.30 | $0.17 |
| Output price per 1M tokens | $1.20 | $0.57 |
| Cached input price per 1M tokens | $0.0060 | $0.06 |
| OpenCode tokens: past 2 months | 140T | 0 |
| Share of all tokens | 16% | 0% |
| Rank by tokens last week | #2 | - |
| Unique users | 1.2M | 1.1K |
| Weekly retention | 58.4% | 33.3% |

Model pages: [DeepSeek V4.1 Flash](https://opencode.ai/data/deepseek/deepseek-v4-1-flash.md), [Hy3 preview](https://opencode.ai/data/tencent/hy3-preview.md)

## Methodology

- Updates: Aggregated every hour. Days and weeks use UTC.
- Tokens: Input, output, reasoning, and cached tokens for each request.
- Users and sessions: Approximate counts of distinct users and OpenCode sessions.
- Cost: Session cost is the average cost per OpenCode session. Token prices are list prices from the OpenCode model catalog.
- Retention: The share of a model's users in one week who use it again the next week.
- Citation: Cite OpenCode Data (opencode.ai/data) with the update time shown at the top of the page.