Changelog
What changed in each release of the API.
0.7.2
Released 2026-10-13. reasoning_effort in
POST /v1/chat/completions, with a default
effort per model class: high for indicia/sterk
and medium for indicia/snel. Claude answers now report
usage.completion_tokens_details.reasoning_tokens. Every model in the
ChatGPT model list is a direct model, including
GPT-6.1 Sol (gpt-6.1-sol), which indicia/sterk now tries after
gpt-6-astra.
0.7.1
Released 2026-10-13. These docs moved to a public site, with guides for using the
API from your coding tool. Dated Claude models are
listed without their date in GET /v1/models.
0.7.0
Released 2026-10-13. x-ratelimit-* headers on responses to service tokens,
GET /v1/usage, PDF and text files in
POST /v1/chat/completions, direct ChatGPT and Claude models in
POST /v1/chat/completions, and readable names in
GET /v1/models.
0.6.1
Released 2026-10-13. The base URL is now https://code.test.indicia-ai.com.
0.6.0
Released 2026-10-13. First release of the developer API: service tokens with
per-token limits, POST /v1/messages and POST /v1/messages/count_tokens,
and images in POST /v1/chat/completions.