Free tier
$10 of API credit, on us.
Try GLM 5.3 Flash through above.dev on a real workload. No card, no trial clock, and the same per-token rates paying customers get, so the credit stretches a long way with prompt caching.
What you get
- $10.00 balance on a new above.dev account, ready the moment you click the link in your email.
- GLM 5.3 Flash, a 1M-context reasoning model with tool use and structured output, served from a dedicated endpoint.
- Paid rates, not trial rates. $0.165 per million input tokens, $0.032 for cache hits, $0.55 output.
- Prompt caching passed through. Coding agents resend the same context every turn. Our testers see cache hit rates above 90%, which is why $10 lasts a couple of weeks of agent use.
- An OpenAI-compatible API. Point any SDK or agent at the base URL below and go.
How far does $10 go?
Roughly, assuming the cache hit rates we see from real coding-agent traffic.
| Workload | Approx. credit used |
|---|---|
| A chat message with a short prompt | under $0.001 |
| An hour of a coding agent working in a large repo | $0.20 to $0.50 |
| A full day of heavy agent use | $0.75 to $1.50 |
Quick start
Your key arrives by email. Then:
curl https://api.above.dev/v1/messages \
-H "Authorization: Bearer $ABOVE_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"glm-5.3-flash-modal","messages":[{"role":"user","content":"Hello"}]}'
For coding agents and SDKs, set the base URL to https://api.above.dev/v1 and the model to glm-5.3-flash-modal. Full details in the docs.
The fine print
- One free grant per email, for new accounts only. Limited to the first 100 accounts.
- The free key is limited to GLM 5.3 Flash. Buy any credit pack and every model in the catalog unlocks on the same key.
- Disposable email addresses are not accepted.
- The credit does not expire, but the free endpoint may be retired when the promotion ends. We will email you first.
- Standard terms and privacy policy apply.