feat:Updated Deep Infra models #135
Conversation
WalkthroughThe pull request introduces updates to the DeepInfra model integration, focusing on pricing adjustments for several AI models and the addition of a new model, QVQ-72B-Preview. The changes span multiple files within the DeepInfra provider implementation, including model identifiers, model provider configuration, and predefined model classes. The modifications primarily involve updating token costs for existing models and expanding the model catalog with a new entry. Changes
Sequence DiagramsequenceDiagram
participant Provider as DeepInfraProvider
participant ModelIds as DeepInfraModelIds
participant Models as Predefined Models
Provider->>ModelIds: Retrieve Model Identifier
ModelIds-->>Provider: Return Model ID (e.g., QVQ-72B-Preview)
Provider->>Models: Initialize Model with Provider
Models-->>Provider: Create Model Instance
Poem
Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media? 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
CodeRabbit Configuration File (
|
There was a problem hiding this comment.
Actionable comments posted: 0
🧹 Nitpick comments (1)
src/DeepInfra/src/DeepInfraModelProvider.cs (1)
21-21: Llama31Nemotron70BInstruct cost updates.Prompt and completion cost are now
1.2E-07/3E-07. If you intend to reflect a major discount, consider highlighting it in the documentation or release notes.
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (3)
src/DeepInfra/src/DeepInfraModelIds.cs(7 hunks)src/DeepInfra/src/DeepInfraModelProvider.cs(3 hunks)src/DeepInfra/src/Predefined/AllModels.cs(1 hunks)
🔇 Additional comments (15)
src/DeepInfra/src/DeepInfraModelProvider.cs (7)
13-13: Decreased pricing for Llama3370BInstructTurbo looks consistent.
The updated prompt and completion costs (1.2E-07 / 3E-07) match the recently announced pricing changes.
16-16: Updated pricing for MetaLlama31405BInstruct.
Good to see an explicit reduction to 8.000000000000001E-07 for both prompt and completion tokens, aligning with your new cost structure.
19-19: Pricing for MetaLlama3170BInstructTurbo updated.
The cost adjustments (1.2E-07 / 3E-07) align well with the new pricing strategy.
20-20: Qwen25Coder32BInstruct cost changes acknowledged.
Reduced costs (7E-08 / 1.6E-07) appear consistent with the file-level updates.
32-32: Hermes3Llama31405B pricing updated successfully.
The new 8.000000000000001E-07 rates are consistent with the rest of the updated cost structure.
34-34: New Qvq72BPreview entry recognized.
Thanks for adding the new model (Qvq72BPreview) at line 34. This maintains the dictionary’s consistent structure and provides a broader model offering.
60-60: Llama323BInstruct adjusted prompt and completion cost.
The new cost values 2E-08 / 2E-08 keep it consistent with the minimal cost tier. Confirm that downstream references have been tested for correctness.
src/DeepInfra/src/Predefined/AllModels.cs (1)
113-117: New Qvq72BPreviewModel class addition looks good.
The class inherits correctly from DeepInfraModel and references the proper enum ID. Good job keeping consistency with existing patterns and documentation comments.
src/DeepInfra/src/DeepInfraModelIds.cs (7)
24-25: Llama3370BInstructTurbo cost reduction confirmed.
Prompt and completion costs updated to $0.12/MTok. This is consistent with your downward pricing strategy for advanced models.
57-58: MetaLlama31405BInstruct cost changes.
Both prompt and completion costs are now $0.8/MTok. Ensure the new rate is appropriately documented in marketing or usage instructions.
90-91: MetaLlama3170BInstructTurbo cost lowered.
Good to see the costs reduced to $0.12/MTok each. This is a straightforward, consistent update.
101-102: Qwen25Coder32BInstruct pricing updated to $0.07/MTok.
Matches the number in DeepInfraModelProvider.cs.
112-113: Llama31Nemotron70BInstruct rates updated.
Now $0.12/MTok. Make sure your upstream billing or usage-tracking system reflects these changes.
233-234: Hermes3Llama31405B prompt and completion cost.
Updated to $0.8/MTok. Please confirm usage references for consistency.
251-261: QVQ-72B-Preview model addition.
The newly introduced model entry includes proper metadata and pricing ($0.25/MTok). Great job documenting its unique multimodal benchmarks.
Created by Github Actions
Summary by CodeRabbit
New Features
QVQ-72B-Previewto the available model listPricing Updates