$ diff hy4-preview gpt-5-6-sol
Hy4 preview vs GPT-5.6 Sol
Values come from Hy4 preview and GPT-5.6 Sol, 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
- 1.1M
- weights
- closed
- $/M in
- $4
- $/M out
- $20
What actually differs
- Context window
- GPT-5.6 Sol takes 1.1M against 1M — modestly more room in a single request.
- Input price
- Hy4 preview at $0.834/M against $4/M — 4.8× cheaper to feed.
- Output price
- Hy4 preview at $2.501/M against $20/M — 8.0× 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); GPT-5.6 Sol is API-only. That decides self-hosting, air-gapped deployment and fine-tuning before any capability question does.
- Recency
- Hy4 preview shipped 50 days after GPT-5.6 Sol (2026-08-28 vs 2026-07-09).
Full spec
| Attribute | Hy4 preview | GPT-5.6 Sol |
|---|---|---|
| Developer | Tencent | OpenAI |
| Released | 2026-08-28 | 2026-07-09 |
| Context window | over 1,000,000 | 1.05M tokens |
| Pricing | $0.834/M input · $2.501/M output | $4/M input · $20/M output (Sol, promotional from 2026-08-21 through at least 2026-11-21; list $5/M · $30/M; family tiers below) |
| License | Apache-2.0 | proprietary (API-only; no weight release) |
| Availability | Hugging Face (BF16 + FP8), ModelScope, GitCode, CNB, Tencent Cloud TokenHub, OpenRouter, WorkBuddy, CodeBuddy, Yuanbao, ima | ChatGPT (subscription), API |