Repository navigation
Record native and public provider facts with regular prices - #12
Merged
Merged
Conversation
softmarshmallow
changed the base branch from
chore/provider-endpoint-schema
to
main
October 13, 2026 12:39
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Keep all 61 existing models and the rig-check operation, and record native/API-provider facts independently. Exact IDs, limits, capabilities and regular prices are not assumed to match between providers.
Final coverage
Corrected and qualified facts
The original calculator/vision/native-parameter work remains included: 79 GPT Image examples, of which 69 have sourced token estimates and ten fal 1920×1080 rows retain quoted USD only because those dimensions violate the 16px rule. Partial native parameter metadata is not a complete provider request schema.
The coverage survey corrects OpenRouter Image 2 text/image/output token rates, 2.5 transparency support, Seedance video-token SKUs, Gemini Embedding 2 input meters, and missing Muse/Lyria/Tripo/provider aliases. Vercel Gemini 3.7/3.8 regular list rates are explicitly $1.50 input/$7.50 output/$0.15 cached input per million tokens; no reverse-discount inference. Provider-specific bands, rounding, regions and optional charges remain qualified.
Native GPT-5.6 Sol, fal Tripo P2, native Tencent Hunyuan and native Bria have no invented regular price where the inspected source does not establish one. OpenRouter Muse's missing image-specific metadata does not imply parameter parity. No promotional schedules, negotiated rates or consumer routing choices are published.
Final publication review: Grida's Gemini 3.7/3.8 Flash bindings retain their verified contracts but omit prices because a regular, non-promotional Grida schedule has not been independently established. Another provider's rates do not establish Grida's advertised rates. Other Grida service rates match inspected public discovery.
Sources and validation
Every changed binding carries first-party sources; qualifications are consolidated in
catalogue/PROVENANCE.md. Principal sources include OpenAI image generation, vision and model pages; Claude pricing and vision; Google pricing and media-resolution; Vercel discovery and individual model pages; OpenRouter discovery and modality/endpoint feeds; exact fal/Tripo API pages; BytePlus pricing; BFL pricing; the native Recraft/xAI/Alibaba/Tencent/Bria pages recorded on their bindings; Grida operation schemas; and official code/model cards.Generated catalogue/schema/invariants, 276 checker/release tests, pinned formatter and skill validation pass. Independent review completed. These are documentation/schema and free checks, not paid behavioral qualification.
Schema #11 was reviewed and merged first. This PR is now based on main; catalogue publication does not change any serving service.