Skip to content

feat(providers): refresh model catalogs and support custom routing - #7851

Merged
waleedlatif1 merged 2 commits into
stagingfrom
codex/cloud-model-catalog
Sep 15, 2026
Merged

feat(providers): refresh model catalogs and support custom routing#7851
waleedlatif1 merged 2 commits into
stagingfrom
codex/cloud-model-catalog

Conversation

@waleedlatif1

@waleedlatif1 waleedlatif1 commented Sep 15, 2026

Copy link
Copy Markdown
Collaborator

Summary

  • Add 49 current model IDs across Azure OpenAI/Anthropic, Bedrock, Vertex, OpenAI, Gemini, DeepSeek, Groq, Cerebras, Mistral, NVIDIA, and Sakana. Verify current families on the remaining providers and retain live discovery for Ollama Cloud and other gateways.
  • Route custom namespaced IDs consistently through workflow validation, credential fields, key selection, and inference. Preserve model-name casing and prevent discovered local names from overriding an explicit cloud provider.
  • Correct documented pricing tiers and parameter support; preserve Bedrock reasoning blocks, Gemini tool-call IDs, and provider-specific tool/JSON behavior. Keep defaults, recommendations, and the Fireworks automatic-routing pool unchanged.
Verification sources and limits

Model IDs, new numeric metadata, and request controls were checked against live provider documentation, with secondary pricing cross-checks. Sources include Azure models, Bedrock model cards, Vertex models, OpenAI models, Gemini models, DeepSeek pricing, Groq models, Cerebras models, Mistral models, Sakana pricing, NVIDIA Lightning, Kimi tool choice, and Ollama Cloud discovery.

Prices use the documented standard region/tier or current promotion; DeepSeek uses peak rates. Optional undocumented limits and controls are omitted. Bedrock Fable 5.1 remains excluded because native structured-output support could not be verified and the schema-with-tools fallback requires forced tool selection. This refresh does not certify historical aliases' variable prices or every deployment's availability. Tests verify request payloads and regression behavior; paid vendor inference was not performed.

Type of Change

  • Feature
  • Bug fix

Testing

  • 2,856 provider/discovery/integration-focused tests passed; final namespace and credential regressions passed (545 tests).
  • Full repository tests passed using CI's dummy test environment: 19 workspace tasks, including 51,570 app tests (247 skipped).
  • All-workspace type-check, lint, all 46 repository audits, block-registry audit, and generated documentation checks.
  • Independent parallel review of routing, credentials, model metadata, and provider implementations.

Checklist

  • Code follows project style guidelines
  • Self-reviewed my changes
  • Tests added/updated and passing
  • No new warnings introduced
  • I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)

@vercel

vercel Bot commented Sep 15, 2026 •
edited
Loading

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

1 Skipped Deployment
Project Deployment Actions Updated
docs Skipped Skipped Sep 15, 2026 4:41am UTC

Request Review

@greptile-apps

greptile-apps Bot commented Sep 15, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

RetriggerConfidence Score: 5/5

The PR appears safe to merge; the prior catalog metadata concern is fixed and no actionable new regression remains.

Summary

This PR refreshes provider model catalogs and makes custom namespaced model IDs route consistently through validation, credential selection, and provider inference.

  • Adds and updates model metadata across supported cloud providers.
  • Supports uncataloged namespaced deployments while preserving model-name casing.
  • Prevents dynamic discovery lists from overriding the explicitly routed provider.
  • Preserves Bedrock reasoning content across tool turns and Gemini tool-call identifiers.
  • Adds regression coverage for routing, credentials, inference payloads, and discovery behavior.

Diagram

 
%%{init: {'theme': 'neutral'}}%%
flowchart LR
    A[Namespaced model ID] --> B[Validate known or custom ID]
    B --> C[Resolve provider from namespace]
    C --> D[Show provider credential fields]
    C --> E[Select provider-specific key]
    E --> F[Execute provider request]
    F --> G[Strip prefix case-insensitively]
    G --> H[Preserve model ID casing on wire]
Loading

Reviews (2) · Last reviewed commit: "fix(providers): add verified Qwen 3.8 re..."

greptile-apps[bot]
greptile-apps Bot reviewed Sep 15, 2026
View reviewed changes
Comment thread apps/sim/providers/models.ts
Comment thread apps/sim/providers/models.ts
@waleedlatif1

waleedlatif1 commented Sep 15, 2026

Copy link
Copy Markdown
Collaborator Author

@greptile

@waleedlatif1

waleedlatif1 commented Sep 15, 2026

Copy link
Copy Markdown
Collaborator Author

@cubic-dev-ai review this PR

@cubic-dev-ai

cubic-dev-ai Bot commented Sep 15, 2026

Copy link
Copy Markdown
Contributor

@cubic-dev-ai review this PR

@waleedlatif1 I have started the AI code review. It will take a few minutes to complete.

@waleedlatif1
waleedlatif1 merged commit e94f069 into staging Sep 15, 2026
33 checks passed
@waleedlatif1
waleedlatif1 deleted the codex/cloud-model-catalog branch September 15, 2026 04:45
@waleedlatif1 waleedlatif1 mentioned this pull request Sep 15, 2026
Merged
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@greptile-apps greptile-apps[bot] greptile-apps[bot] left review comments

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@waleedlatif1