Tokenizer V5
Nepali BPE tokenization and decoding
/api/v1/tokenizer
Supernova API turns the published Supernova engines into simple, observable endpoints for tokenization, language detection, and text normalization.
POST /api/v1/tokenizer/encode
// actual engine output
{
"tokens": [881, 3490, 389, 376, 404, 378, 402, 3506, 410],
"count": 9,
"request_id": "req_01J..."
}
Three focused engines
Each endpoint is designed to return the underlying package result, with request IDs, validation, and usage accounting around it.
Nepali BPE tokenization and decoding
/api/v1/tokenizer
Language, script, and confidence signals
/api/v1/language/detect
Canonical Nepali text normalization
/api/v1/text/normalize
Developer first
Use API keys for every request. Credits, rate limits, and usage logs are enforced on the server; your key is only shown at creation.
Explorer accounts start with a one-time 100 SCC allowance.
Keys are hashed at rest and can be revoked at any time.
Every successful request is metered in the persistent ledger.
curl -X POST https://api.supernova.dev/api/v1/tokenizer/encode \
-H "Authorization: Bearer snv_live_..." \
-H "Content-Type: application/json" \
-d '{"text":"नेपाल राम्रो छ।"}'Supernova Contribution Credits
SCC is an internal access unit, not cash or cryptocurrency. Earn credits by contributing code, docs, testing, infrastructure, domains, or approved service upgrades. Every award is reviewed by an administrator and recorded in an auditable ledger.