Add support for deepseek-r1-0528 model#168
Conversation
- Added deepseek-r1-0528 to MODEL_CONFIG with 130k token limit - Added to all pricing plan tiers (excluded from Free/Starter, included in Pro/Max/Team) - Following same pattern as existing deepseek-r1-70b model 🤖 Generated with [Claude Code](https://claude.ai/code) Co-authored-by: Anthony <AnthonyRonning@users.noreply.github.com>
WalkthroughA new DeepSeek model ("deepseek-r1-0528") with a 130,000 token limit was added to the model configuration and pricing plan feature lists. The model is marked as requiring a Pro subscription. Additionally, minor formatting changes were made to an error handler in the chat route, without affecting functionality. Changes
Sequence Diagram(s)sequenceDiagram
participant User
participant UI (ModelSelector)
participant PricingConfig
User->>UI (ModelSelector): Request available models
UI (ModelSelector)->>UI (ModelSelector): Show "DeepSeek R1 0528" if Pro
User->>PricingConfig: View plan features
PricingConfig->>User: Show "DeepSeek R1 0528" availability by plan
Estimated code review effort🎯 1 (Trivial) | ⏱️ ~3 minutes Assessment against linked issues
Assessment against linked issues: Out-of-scope changesNo out-of-scope changes found. Possibly related PRs
Poem
📜 Recent review detailsConfiguration used: CodeRabbit UI 📒 Files selected for processing (1)
🚧 Files skipped from review as they are similar to previous changes (1)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (5)
✨ Finishing Touches
🧪 Generate unit tests
🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed help? Create a ticket on our support page for assistance with any issues or questions. 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 (
|
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
There was a problem hiding this comment.
Greptile Summary
This PR adds support for the DeepSeek R1 0528 model, a new AI model with a 130k context window. The implementation follows the established architectural pattern used for adding AI models to the Maple platform.
The core change is in ModelSelector.tsx where the new model is added to the MODEL_CONFIG object with a displayName of "DeepSeek R1 0528", a "Pro" badge indicating it requires a paid subscription, requiresPro: true for access control, and a tokenLimit of 130,000 tokens (double the existing deepseek-r1-70b model's 64k limit).
The model is then integrated into the pricing tier system in pricingConfig.tsx by adding "DeepSeek R1 0528" as a feature across all pricing plans. It follows the same access pattern as the existing DeepSeek R1 70B model - excluded from Free and Starter tiers but included in Pro, Max, and Team tiers. This maintains the business model of offering advanced AI models as premium features.
A minor formatting improvement was also made to the error handling in the chat component, reformatting a catch block callback from single-line to multi-line for better readability without changing functionality.
The changes integrate seamlessly with the existing model selection system, pricing infrastructure, and access control mechanisms without requiring any database migrations or breaking changes.
Confidence score: 5/5
- This PR is extremely safe to merge with no apparent issues or risks
- The implementation follows established patterns perfectly, uses consistent naming, and maintains proper access controls across all affected files
- No files need additional attention - all changes are straightforward and follow existing conventions
3 files reviewed, no comments
|
@claude let's add |
|
Claude finished @AnthonyRonning's task —— View job ✅ Task CompleteI've successfully updated the DeepSeek model name to include "671B" as requested. Changes Made:
Todo List:
All formatting, linting, and build checks passed successfully. The changes have been committed and pushed to the PR branch. |
- Updated displayName in ModelSelector.tsx from 'DeepSeek R1 0528' to 'DeepSeek R1 0528 671B' - Updated all pricing config references to match the new model name Co-authored-by: Anthony <AnthonyRonning@users.noreply.github.com>
There was a problem hiding this comment.
Greptile Summary
This PR adds support for the new DeepSeek R1 0528 671B model to the Maple AI chat application. The changes integrate a new AI model variant that offers significantly enhanced capabilities with a 130k token context window (compared to 64k for the existing deepseek-r1-70b model) and 671 billion parameters.
The implementation follows the established architectural pattern for premium models in the codebase. In ModelSelector.tsx, the new model is configured with requiresPro: true and a 130k token limit, ensuring it's restricted to paid subscription tiers. The pricingConfig.tsx file is updated across all pricing plans to maintain consistency - the model is explicitly excluded from Free and Starter tiers while being included as a feature for Pro, Max, and Team plans.
The naming convention aligns with existing models by including the parameter count ('671B') in the display name, helping users understand the model's scale and capabilities. This change expands the AI model offerings available to premium users while maintaining the existing business model structure that gates advanced features behind paid subscriptions.
Confidence score: 5/5
- This PR is extremely safe to merge with no apparent issues or risks.
- The implementation perfectly follows established patterns, maintains consistency across all configuration files, and includes proper access controls for the premium model.
- No files need additional attention - both changed files are properly updated and consistent with each other.
2 files reviewed, no comments
Deploying maple with
|
| Latest commit: |
97547f2
|
| Status: | ✅ Deploy successful! |
| Preview URL: | https://1a81c690.maple-ca8.pages.dev |
| Branch Preview URL: | https://claude-issue-167-20250730-15.maple-ca8.pages.dev |
…epSeek R1 671B - Changed badge property from string to string array for all models - Added Beta tag to DeepSeek R1 0528 671B model alongside Pro tag - Added yellow-orange gradient styling for Beta badges - Increased model selector dropdown width from w-64 to w-72 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>
There was a problem hiding this comment.
Greptile Summary
This PR adds support for DeepSeek's new R1 0528 model, a 671B parameter AI model with a 130k context window. The change introduces the model to the MODEL_CONFIG with appropriate configuration and makes it available to Pro, Max, and Team plan users while excluding it from Free and Starter tiers.
The implementation required a significant refactoring of the badge system in the ModelSelector component. Previously, models could only display a single badge, but the new model needs both 'Pro' (indicating tier requirement) and 'Beta' (indicating experimental status) badges. This led to changing the type definition from a single badge property to a badges array and updating the rendering logic accordingly.
The change follows the established pattern in the codebase where premium models are restricted to paying customers. The new model is positioned as a more powerful alternative to the existing deepseek-r1-70b model, offering double the context window (130k vs 64k tokens). The dropdown width was also increased from w-64 to w-72 to accommodate the visual space needed for multiple badges.
This enhancement integrates seamlessly with the existing pricing and model selection infrastructure, maintaining consistency with how other premium models are handled throughout the application.
Confidence score: 4/5
• This PR appears safe to merge with well-structured changes that follow existing patterns
• The refactoring is clean and the new model configuration follows established conventions, though the badge system change is significant
• The ModelSelector.tsx file should be tested thoroughly to ensure the badge refactoring doesn't break existing model displays
1 file reviewed, no comments
1 participant
Add support for the new deepseek-r1-0528 model with 130k context window.
Resolves #167
Generated with Claude Code
Summary by CodeRabbit
New Features
Style