feat(together-ai/Salesforce/Llama-Rank-V1): add new models [bot]

[models-bot]

Auto-generated by model-addition-agent for together-ai/Salesforce/Llama-Rank-V1.

---

Note

Low Risk
Metadata-only addition of a new model catalog file; no runtime or auth logic changes.

Overview
Adds a new Together AI provider definition for Salesforce/Llama-Rank-V1 as an active rerank model (text input, 8192 context window).

The entry includes per-token pricing (1e-7 input/output), supportedModes: rerank, strips reasoning_effort, and links Together’s LlamaRank / rerank API docs as sources.

^{Reviewed by Cursor Bugbot for commit 9265af2. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Non-chat output token pricing

Medium Severity

output_cost_per_token is set to the same rate as input for this rerank model. Rerank APIs typically bill query/document tokens only; other rerank entries use zero output token cost, and Together embedding configs omit output pricing entirely.

 

^{Reviewed by Cursor Bugbot for commit 7102ff3. Configure here.}

[github-actions]

/test-models

[harshiv-26]

Gateway test results

• Total: 1
• Passed: 0
• Failed: 0
• Validation failed: 0
• Errored: 0
• Skipped: 1
• Success rate: 0.0%

Provider	Model	Scenarios
together-ai	Salesforce/Llama-Rank-V1	skipped: skip-check

Skipped (1)

together-ai/Salesforce/Llama-Rank-V1 — skip-check (skipped)

Skip reason

Provisioned model

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

There are 2 total unresolved issues (including 1 from previous review).

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 9265af2. Configure here.}

[cursor]

Case colliding model paths

Medium Severity

This change adds a full config at providers/together-ai/Salesforce/Llama-Rank-V1.yaml while providers/together-ai/salesforce/Llama-Rank-V1.yaml already declares the same model value with isDeprecated: true and mode: unknown. On case-insensitive filesystems the two org folders collapse, so only one Llama-Rank-V1.yaml can exist and checkout may not match the Linux tree; on Linux, two YAMLs still advertise conflicting metadata for one model ID.

 

^{Reviewed by Cursor Bugbot for commit 9265af2. Configure here.}