feat: add Helicone gateway endpoint for Gemini 2.5 Flash

[colegottdank]

Summary

• Adds gemini-2.5-flash:helicone endpoint with providerModelId pa/gmn-2.5-fls-lt-pw-06-17
• Pricing matches Google's rates ($0.30/1M input, $2.50/1M output)
• Enables PTB (pass-through billing) for the endpoint

Test plan

• Deploy and test with: curl -s "https://ai-gateway.helicone.ai/v1/chat/completions" -H "Authorization: Bearer $API_KEY" -H "Content-Type: application/json" -d '{"model": "gemini-2.5-flash", "messages": [{"role": "user", "content": "Hello"}]}'
• Verify response headers show helicone-provider: helicone

🤖 Generated with Claude Code

[vercel]

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

Project	Deployment	Actions	Updated (UTC)
helicone	Ready	Preview, Comment	Feb 18, 2026 8:57pm
helicone-bifrost	Building	Preview, Comment	Feb 18, 2026 8:57pm
helicone-eu	Ready	Preview, Comment	Feb 18, 2026 8:57pm

[supabase]

This pull request has been ignored for the connected project bolqqmqbrciybnypvklh because there are no changes detected in supbase/supabase directory. You can change this behaviour in Project Integrations Settings ↗︎.

---

Preview Branches by Supabase.
Learn more about Supabase Branching ↗︎.

[supabase]

This pull request has been ignored for the connected project lmahfbbnchpworytrrqk because there are no changes detected in .supabase/supabase directory. You can change this behaviour in Project Integrations Settings ↗︎.

---

Preview Branches by Supabase.
Learn more about Supabase Branching ↗︎.

[greptile-apps]

Your free trial has ended. If you'd like to continue receiving code reviews, you can add a payment method here.

[claude]

Claude finished @colegottdank's task —— View job

---

PR Review: Gemini 2.5 Flash Helicone Endpoint

Todo List:

• Read changed files to understand implementation
• Analyze code for bugs, security issues, and best practices
• Check pricing configuration and model mapping
• Review test snapshot changes
• Provide comprehensive feedback with confidence score

---

Score: 9/10 - Excellent implementation with consistent patterns and correct configuration

Analysis

This PR adds a new Helicone gateway endpoint for Gemini 2.5 Flash, following established patterns in the codebase. The implementation is clean and consistent with other provider endpoints.

Code Quality Assessment

✅ Strengths:

1. Consistent Pattern Matching: The implementation follows the exact same structure as other Helicone endpoints (o1:helicone, claude-*:helicone, etc.)

2. Correct Pricing Configuration:

   • Input: $0.30/1M tokens (0.0000003) matches Google's official rates
   • Output: $2.50/1M tokens (0.0000025) matches Google's official rates
   • No unnecessary audio pricing included (unlike google-ai-studio/vertex variants)

3. Proper Model ID: Uses the pa/ prefix pattern consistent with other Helicone endpoints:

   providerModelId: "pa/gmn-2.5-fls-lt-pw-06-17"

4. Complete Parameter Support: Includes all reasoning-related parameters (include_reasoning, reasoning, structured_outputs) that other Gemini variants support

5. Correct Configuration:

   • ptbEnabled: true ✅ (enables pass-through billing as mentioned in PR)
   • contextLength: 1_048_576 ✅ (matches spec)
   • maxCompletionTokens: 65_535 ✅ (matches spec)
   • endpointConfigs: { "*": {} } ✅ (standard pattern)

6. Test Snapshot Update: The generated snapshot correctly reflects the new endpoint configuration

🔍 Minor Observations:

1. Missing responseFormat: Unlike the google-ai-studio and vertex variants, the Helicone endpoint doesn't specify responseFormat: "GOOGLE". This appears intentional since other Helicone endpoints also omit this field.

2. No crossRegion field: Also consistent with other Helicone endpoints that don't specify regional configurations.

Security & Best Practices

• No security concerns identified
• Follows established naming conventions
• Uses proper TypeScript typing with satisfies constraint
• No hardcoded secrets or sensitive data

Suggestions Summary:

No critical issues found. The implementation is solid and follows all established patterns correctly.

Minor Enhancement (Optional):
Consider adding a brief comment explaining the pa/ prefix convention for future maintainers:

// Helicone provider model IDs use 'pa/' prefix for internal routing
providerModelId: "pa/gmn-2.5-fls-lt-pw-06-17",

Test Plan Verification

The proposed curl test command looks appropriate and will properly verify the endpoint functionality with the expected helicone-provider: helicone header.

Final Recommendation: ✅ Approve and merge - This is a well-implemented feature addition that maintains consistency with the existing codebase.