770B mode, 49B active, 1M context. cmd update available in GOAT and above plans. Our early internal benchmarks show strong perf yet cheaper per task compared to GLM 5.3 Flash.
Y'all share how it goes. 🐐
Tencent Hy4 Preview at a glance
Agentic coding & sustained multi-step tool use.
| Spec | Tencent Hy4 Preview |
|---|---|
| Model ID | tencent/hy4-preview |
| Context window | 1.05M tokens |
| Released | August 27, 2026 |
| Input | $0.83 per 1M tokens |
| Output | $2.50 per 1M tokens |
| Cache read | $0.04 per 1M tokens |
| Effective input in an agent loop | $0.28 per 1M tokens |
| Plans | Go and above |
Specs, pricing and plans are from the Tencent Hy4 Preview model page as of September 28, 2026, including any deal running that day, so they can differ from the launch pricing above. Intelligence and Coding Index scores are from Artificial Analysis.
Use Tencent Hy4 Preview
1cmd --model tencent/hy4-previewOr type /model in a session and pick it. You can switch mid-session without losing context.
What real work costs on Tencent Hy4 Preview
| Task | Tokens (in · out) | Cost |
|---|---|---|
| Quick lookup / one-liner | 8K · 1K | $0.0035 |
| Review a 500-line PR | 60K · 4K | $0.03 |
| Fix a bug (agent loop) | 180K · 12K | $0.07 |
| Refactor a module | 320K · 20K | $0.13 |
| Full-repo agent run | 900K · 45K | $0.31 |
Try Command Code
The best coding agent for open models, in your terminal or on your desktop.
1npm i -g command-codeDesktop App · Docs · X · Discord
