We post product updates here and announce breaking API changes in advance. Subscribe via the RSS link below or follow @perchy_ai for the short version.
Three new long-context models added to the catalog. M2.7 is in the General Purpose lane (1M context, multi-modal). V4 Flash drops in as the cheapest per-token option for production chat.
The HTTP API at api.perchy.ai/v1 is now stable. Existing OpenAI client libraries work by changing the base URL and key — no SDK install required.
One-time plan purchases and usage extensions stay on perchy.ai. Signed webhooks fulfill the order and update the console immediately.
New /account/usage view: per-day tokens, per-model spend, and recent calls. Backed by a stable /v1/usage endpoint for programmatic access.
Hosts can finish KYC and link payouts directly inside the console under Settings → Payouts. First payout cycle for new hosts is the Friday after KYC clears.