$ diff hy4-preview claude-opus-5
Hy4 preview vs Claude Opus 5
Values come from Hy4 preview and Claude Opus 5, where each is cited to its source. This page states no benchmark result and ranks neither model — it puts two published specifications next to each other. Where a lab has not published a figure, the row says so rather than guessing.
TencentHy4 preview
- context
- 1M
- weights
- open
- $/M in
- $0.834
- $/M out
- $2.501
- context
- 1M
- weights
- closed
- $/M in
- $5
- $/M out
- $25
What actually differs
- Context window
- Identical — both accept 1M tokens, so context length is not a reason to pick either.
- Input price
- Hy4 preview at $0.834/M against $5/M — 6.0× cheaper to feed.
- Output price
- Hy4 preview at $2.501/M against $25/M — 10.0× cheaper to generate. Output dominates the bill on most agentic workloads, where the model writes far more than it reads.
- Weights
- Hy4 preview publishes weights (Apache-2.0); Claude Opus 5 is API-only. That decides self-hosting, air-gapped deployment and fine-tuning before any capability question does.
- Recency
- Hy4 preview shipped 35 days after Claude Opus 5 (2026-08-28 vs 2026-07-24).
Full spec
| Attribute | Hy4 preview | Claude Opus 5 |
|---|---|---|
| Developer | Tencent | Anthropic |
| Released | 2026-08-28 | 2026-07-24 |
| Context window | over 1,000,000 | 1M tokens |
| Pricing | $0.834/M input · $2.501/M output | $5/M input · $25/M output (standard); $10/M · $50/M (Fast mode, research preview) |
| License | Apache-2.0 | proprietary |
| Availability | Hugging Face (BF16 + FP8), ModelScope, GitCode, CNB, Tencent Cloud TokenHub, OpenRouter, WorkBuddy, CodeBuddy, Yuanbao, ima | Claude.ai, API (`claude-opus-5`), Bedrock (`anthropic.claude-opus-53`), Vertex AI, Microsoft Foundry, Claude Code, Cowork |