Repository navigation
feat(perplexity-ai): update model YAMLs [bot] - #5659
Open
models-bot[bot] wants to merge 14 commits into
Open
models-bot[bot] wants to merge 14 commits into
models-bot[bot] wants to merge 14 commits into
Conversation
Introduce reviewed canonical_model mappings and optional Artificial Analysis ids for chat/completion models, with proposal/apply/match/check scripts, CUE schema support, docs, and identity validation.
Wire check_model_identity and changed-file Artificial Analysis verification into validate.yml. Identity tags are reviewed on model YAML in PRs, so remove committed canonical_aliases.yaml and canonical_external_ids.yaml. Keep canonical_exceptions.yaml and optional bulk apply helpers.
CI was failing because the repository secret is not configured yet. Identity structure checks remain required; live AA checks activate once AA_API_KEY is set.
Neither field is required on chat/completion models. Remove canonical_exceptions.yaml and relax identity validation to check format and consistency only when the fields are present.
Fetch the exact PR base SHA without --depth=1, and fall back to a direct tree diff if three-dot merge-base resolution fails.
Raw SHA refspecs often fail with "couldn't find remote ref" and abort the identity job under set -e. Fetch the base branch, then diff against the PR base SHA when that object is available.
YAML rewrite had rounded cache_creation_input_token_cost from 8.333333333333334e-8 to 8.333333333333333e-8. Revert those values so only identity fields differ from main.
Rebuild each modified model YAML from main and surgically insert canonical_model / external_ids so costs, params, comments, and null style are unchanged.
When both identity keys were inserted at the same line index, external_ids was placed above canonical_model. Insert larger keys first so alphabetical top-level order is preserved.
Contributor
SDK live feature checksResult: failed (1/1 scenarios) Informational — does not block merge.
Failures
|
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.
❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.
Reviewed by Cursor Bugbot for commit 18875cd. Configure here.
| from: 100000 | ||
| output: | ||
| - cost_per_token: 2.5e-6 | ||
| from: 100000 |
There was a problem hiding this comment.
Missing cumulative long-context pricing mode
Medium Severity
The new tiered_pricing block omits pricing_mode, so it defaults to marginal. Perplexity prices Haiku 5.5 by prompt length: once input exceeds 100k tokens, every token uses the long-context rate. Cost calculations will undercharge those requests.
Reviewed by Cursor Bugbot for commit 18875cd. Configure here.
…tity-9a44 Add canonical model names and Artificial Analysis external IDs
…ry Managed models (#5646) * Add explicit reasoning_effort on account OpenAI, Claude and TrueFoundry Managed models * Align TrueFoundry Managed reasoning_effort with provider docs
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
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
4 participants
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.
Auto-generated by poc-agent for provider
perplexity-ai.Note
Medium Risk
Very wide YAML-only metadata rollout with new CI gates; incorrect or inconsistent
external_idscan fail PR checks when the AA secret is configured, but no billing or routing logic is modified in this diff.Overview
Introduces optional foundation-model identity on provider YAML:
canonical_model(<creator>/<foundation-model>) and verifiedexternal_ids(Artificial Analysisid/slug). CUE and generated TypeScript types are extended accordingly; CONTRIBUTING and README document how to set and review these fields.CI and tooling: A new Validate model identity workflow runs
check_model_identity.py(format and consistency within canonical groups) andverify_changed_external_ids.py(live catalog check for changedexternal_idswhenAA_API_KEYis set). Tracked Python scripts underscripts/, npm targets (validate:identity,test:identity), and.gitignoreupdates support this.Data rollout: A large bulk edit adds
canonical_modelacross provider configs (Anthropic, AWS Bedrock/Mantle, AI21, and many others) and attaches matchingexternal_idswhere reviewed. Deployments that share weights reuse the same canonical name and AA pair (e.g. direct Anthropic vs Bedrock regional entries). A few Anthropic entries also gainreasoning_effortparam definitions.Provider pricing and API
modelstrings are unchanged; identity fields are cross-reference metadata only.Reviewed by Cursor Bugbot for commit 22e1852. Bugbot is set up for automated code reviews on this repo. Configure here.