Canopywave deepseek v3.2 speciale

[cynthia-lixinyi]

Ticket

Link to the ticket(s) this pull request addresses.

Component/Service

What part of Helicone does this affect?

• Web (Frontend)
• Jawn (Backend)
• Worker (Proxy)
• Bifrost (Marketing)
• AI Gateway
• Packages
• Infrastructure/Docker
• Documentation

Type of Change

• Bug fix
• New feature
• Breaking change
• Documentation update
• Performance improvement
• Refactoring

Testing

• Added/updated unit tests
• Added/updated integration tests
• Tested locally
• Verified in staging environment
• E2E tests pass (if applicable)

Technical Considerations

• Database migrations included (if needed)
• API changes documented
• Breaking changes noted
• Performance impact assessed
• Security implications reviewed

Dependencies

• No external dependencies added
• Dependencies added and documented
• Environment variables added/modified

Deployment Notes

• No special deployment steps required
• Database migrations need to run
• Environment variable changes required
• Coordination with other teams needed

Context

Why are you making this change?

Screenshots / Demos

Before	After

Misc. Review Notes

[vercel]

@cynthia-lixiny07 is attempting to deploy a commit to the Helicone Team on Vercel.

A member of the Team first needs to authorize it.

[greptile-apps]

Greptile Overview

Greptile Summary

Added support for DeepSeek V3.2 Speciale model through Canopy Wave provider with standard configuration matching other similar integrations.

• Added new model definition for deepseek-v3.2-speciale with 163K context length and 65K max output tokens
• Configured Canopy Wave provider endpoint with fp8 quantization, $0.27/$0.41 per 1M tokens pricing
• Updated test snapshots to reflect the new model (4th DeepSeek V3 variant, bringing Canopy Wave's total to 4 models)

Confidence Score: 5/5

• Safe to merge - clean provider integration following established patterns
• Changes follow exact same pattern as previous Canopy Wave integrations, all configuration values are consistent with similar models, test snapshots properly updated, no breaking changes or logic issues
• No files require special attention

Important Files Changed

File Analysis

Filename	Score	Overview
packages/cost/models/authors/deepseek/deepseek-v3/model.ts	5/5	Added deepseek-v3.2-speciale model configuration with 163K context and 65K max tokens
packages/cost/models/authors/deepseek/deepseek-v3/endpoints.ts	5/5	Added Canopy Wave provider endpoint for deepseek-v3.2-speciale with fp8 quantization and standard pricing
packages/tests/cost/snapshots/registrySnapshots.test.ts.snap	5/5	Updated test snapshots to reflect new model and endpoint configuration

Sequence Diagram

sequenceDiagram
    participant User
    participant Registry
    participant ModelConfig
    participant EndpointConfig
    participant CanopyWave

    User->>Registry: Request deepseek-v3.2-speciale
    Registry->>ModelConfig: Lookup model definition
    ModelConfig-->>Registry: Returns model specs (163K context, 65K max tokens)
    Registry->>EndpointConfig: Lookup provider endpoint
    EndpointConfig-->>Registry: Returns Canopy Wave config (fp8, pricing)
    Registry->>CanopyWave: Route request to inference.canopywave.io
    CanopyWave-->>User: Return model response

Loading

[greptile-apps]

_{2 files reviewed, no comments}

_{Edit Code Review Agent Settings | Greptile}

[juliettech13]

Release date is December 1st, 2025

[juliettech13]

Can you double check this context length? Online I'm seeing the context length is 131_072.

[juliettech13]

^ and this looks like 128_000.