Studio99
Studio99 Indic Typography API
Current version 1.2

Changelog

Every change a developer can observe in the Studio99 Indic Typography API: new endpoints and fields, changed defaults, new limits, fixes. Breaking changes get a new major version and a notice here first.

Versioning: the path stays /api/v1 for additive changes. Removals or changed semantics would ship as /api/v2 with v1 kept running for at least 12 months. See the status page for live health and the OpenAPI spec for the current contract.

v1.2

OpenAPI spec, public benchmark prompts, one name

  • addedOpenAPI 3.1 description served at /developers/openapi.json, generated from the same config the API runs on.
  • addedBenchmark prompt set and rubric downloadable at /developers/benchmark/prompts.json.
  • changedThe API is named "Studio99 Indic Typography API" everywhere; /health and /capabilities now report the same version (1.2) and product name.
v1.2

Library matches for first-party channels

  • changedOn the TypoGen-X engine, first-party channel keys receive up to 4 matching Inkora library artworks in `libraryResults`. Public keys are unaffected.
v1.2

Credits model and output gates (flag-gated)

  • addedCredit accounting: 1 credit per variant returned by /generate, 1 per output from /render, 1 per library download. Usage meter gains `unit`, `creditsCharged` and `license`.
  • addedNew Basic plan (500 credits a month).
  • changedFree plan output becomes a watermarked JPG `preview` with no commercial licence once the credits model is enforced; paid plans keep clean SVG and PNG up to 4096 px.
  • fixedDebits are authoritative: a request can no longer generate past a zero balance.
  • addedNew error code INSUFFICIENT_CREDITS (HTTP 429).
v1.2

TypoGen-X engine on the API

  • addedNew default generation engine TypoGen-X with curated slates. Per-request `engine` override (`typogen-x` or `classic`).
  • addedComposition controls: `align` (center, left, right), `lines` (auto or 1 to 6), `lineGap`, `recipe` (plain, flagship, duo, free) and `seed` for reproducible output.
  • addedPer-result `label`, `archetype` and `lineCount`; response `metadata.lines` shows the line split used.
  • fixedDevanagari shaping: anusvara and other combining marks, word-final halant and ligature rendering corrected.
v1.1

AI-native semantic layer

  • addedGET /capabilities: key-less discovery of languages, formats, vocabularies and endpoints for agents.
  • added`use_case` and `mood` on /generate and as filters on /fonts (vocabulary in /capabilities).
  • added`language` is optional on /generate; the API auto-detects the script and transliterates Latin input.
  • addedMCP server at mcp.studio99.app exposes the same engine to Claude, ChatGPT and other MCP clients.
v1.0

Public launch

  • addedPOST /generate returns per-result `png` (base64) when `format: "png"` is requested, clamped to plan limits.
  • addedPOST /render: exact re-rendering of a phrase in a chosen `fontId`, with `fontSize` and `format`.
  • addedGET /library/{id}/download (signed URL, 5 minutes) and GET /library/{id}/render-svg.
  • changedLanguage values validated: hindi, marathi, gujarati, english, or hi, mr, gu, en.
  • addedSelf-service keys and paid plans at accounts.studio99.app.
v0.9

Private API

  • addedPOST /generate, GET /fonts, GET /library/search, GET /library/{id}, GET /health, X-API-Key authentication, per-plan rate limits. Used by Studio99's own Canva, Adobe Express and Android apps before public launch.