Project access
Issue scoped credentials and organize usage by project, environment, or business unit.
02 / RAYA TOKEN APIRAYA.COM.AI
Local intelligence, consumed by token.
One API gives commercial teams access to RAYA localized models for generation, structured outputs, embeddings, and application workflows.
POSITION
The Token API is how RAYA turns model localization into a usable commercial product: consistent access, measured consumption, and enterprise controls.
QUICKSTART / REQUESTAPI.RAYA.COM.AI/V1
01
02
03
04
05
06
07
08
02
03
04
05
06
07
08
curl https://api.raya.com.ai/v1/responses \
-H "Authorization: Bearer $RAYA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "raya-local",
"input": "Summarize this Indonesian report."
}'RESPONSE200 / STREAMLocalized model outputUsage returned by token
COMMERCIAL ACCESS
Preview Token pricing
Indicative rates in Indonesian Rupiah per one million Tokens. Final rates remain subject to launch validation.
MODEL PROFILEINPUT / 1MOUTPUT / 1M
RAYA Local / CompactRp3,000Rp6,000
RAYA Local / GeneralRp8,000Rp24,000
RAYA Local / ReasonRp10,000Rp40,000
RAYA Embed / MultilingualRp1,500—
PLANNED API CONTRACT
A predictable path to production
AuthenticationBearer project keyScoped by environment
Generation/v1/responsesStreaming + structured output
Embeddings/v1/embeddingsRetrieval + semantic search
UsageInput / output TokensReturned per response
ErrorsConsistent status modelRequest ID for support
VersioningStable model identifiersDocumented migration window
This interface describes the intended developer experience. Final endpoints, SDKs, limits, and availability will be confirmed before public access.
01 / APPROACH
How it works
Model profile
Select the RAYA localized profile that fits language, domain, quality, and workflow needs.
Input and output tokens
Use streaming responses, structured generation, embeddings, and tool-enabled patterns.
Operational visibility
Monitor consumption, latency, errors, limits, and policy events across applications.
02 / SCOPE
What it includes
Interface
Consistent API patterns for generation and embeddings
→Responses
Streaming, structured outputs, and tool-oriented workflows
→Operations
Project keys, usage controls, observability, and limits
→Enterprise
Private endpoints and dedicated model profiles where required
→03 / QUESTIONS
Frequently asked
What is an AI Token?
A token is the unit used to measure text processed and generated by a model. RAYA commercializes access to localized models through input and output Token usage.
Is the API only for Bahasa Indonesia?
No. RAYA models are designed for Indonesian requirements while retaining useful multilingual capabilities from their open-model foundations.
Can enterprise customers use dedicated access?
Dedicated endpoints and private delivery patterns can be configured when workload sensitivity or operating requirements justify them.