Build next-gen apps with hy3-preview
Pricing
TokenLab Price
Input $0.1765 / Output $0.5882
| Official Price | TokenLab Price | Discount | |
|---|---|---|---|
| Input tokens <= 16.38K | Input $0.1817 / Output $0.6055 | Input $0.1817 / Output $0.6055 | - |
| Input tokens 16K-32.77K | Input $0.2422 / Output $0.9689 | Input $0.2422 / Output $0.9689 | - |
| Input tokens > 32K | Input $0.3028 / Output $1.21 | Input $0.3028 / Output $1.21 | - |
| Cache Read | $0.0588 | $0.0588 | - |
One-click test
Sign in once and Web Agent keeps this model, prompt, and request preset for you.
Test hy3-preview in Web Agent with a short request at /v1/chat/completions, then show request body, latency, and response.
API workbench
The default route for production. The code sample below uses this endpoint with the format you pick.
curl https://api.tokenlab.sh/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer sk-xxx" \
-d '{
"model": "hy3-preview",
"messages": [
{"role": "user", "content": "Hello!"}
]
}'Use cases
Best forReasoning
Multi-step reasoning, analysis, and research workflows
Agents and tools
Drive reasoning, support triage, tool calls, and multi-step task flows.
Developer workflows
Generate, review, or debug code without rewiring your stack.
Knowledge assistants
Ship chat, search, and retrieval with predictable cost and behavior.
Side-by-side test
See real response quality, latency, and price before this becomes your production default.
Prompt examples
Write a concise support reply and list the assumptions behind it.
Review this API design and call out the top three integration risks.
Turn a long changelog into release notes a non-engineer would read.
Cost Calculator
FAQ
How much does hy3-preview cost?
On TokenLab, hy3-preview costs Input $0.1765 / Output $0.5882 Per Token. The pricing table above shows the full breakdown.
What is hy3-preview best for?
hy3-preview is a strong fit for Reasoning, Tool Use, JSON Mode. You can call it through TokenLab with one API key.
How do I call the hy3-preview API?
Get a TokenLab API key, then send your request to https://api.tokenlab.sh/v1/chat/completions. The API workbench above has a recommended endpoint and copy-ready code.
Which endpoint should hy3-preview use?
Use https://api.tokenlab.sh/v1/chat/completions as the default for hy3-preview. If a provider-native format is supported, the API workbench shows that endpoint too.
Can I test hy3-preview before integrating it?
Yes. Try in Web Agent opens a ready test for hy3-preview and keeps your prompt after sign-in, so you don’t lose context.
