feat(models): added xai models and updated gemini pricing

[waleedlatif1]

Summary

• added xai models and updated gemini pricing

Type of Change

• New feature

Testing

Tested manually

Checklist

• Code follows project style guidelines
• Self-reviewed my changes
• Tests added/updated and passing
• No new warnings introduced
• I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)

[vercel]

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

1 Skipped Deployment

Project	Deployment	Preview	Comments	Updated (UTC)
docs	Skipped			Dec 2, 2025 10:31pm

[greptile-apps]

Greptile Overview

Greptile Summary

Updated Google Gemini 2.5 model pricing to reflect significant increases (Pro input: 0.15→1.25, Flash input: 0.15→0.3), added new gemini-2.5-flash-lite model, and expanded xAI Grok models with 4 new variants (grok-4-0709, grok-4-1-fast-reasoning, grok-4-1-fast-non-reasoning) plus updated cached input pricing for grok-4-latest.

Major changes:

• Gemini 2.5 Pro pricing increased substantially (8.3x input, 16.7x output)
• Added gemini-2.5-flash-lite with lower pricing (input: 0.1, output: 0.4)
• Added 4 new xAI Grok model variants with 2M context window for fast models
• Updated grok-4-latest cached input pricing from 1.5 to 0.75

Critical issues found:

• grok-4-fast-reasoning and grok-4-fast-non-reasoning models have cachedInput (0.25) priced higher than input (0.2), which is illogical as cached input should cost less

Confidence Score: 2/5

• This PR contains logical errors in pricing configuration that will cause incorrect cost calculations
• The PR has critical pricing logic errors where cachedInput costs more than input for two models (grok-4-fast-reasoning and grok-4-fast-non-reasoning). This contradicts the fundamental principle that cached input should be cheaper than regular input. The Gemini pricing updates and new model additions appear correct, but the xAI cached pricing errors need to be fixed before merging.
• apps/sim/providers/models.ts:839-840 and apps/sim/providers/models.ts:852-853 need immediate correction for cachedInput pricing

Important Files Changed

File Analysis

Filename	Score	Overview
apps/sim/providers/models.ts	2/5	Updated Gemini 2.5 pricing, added gemini-2.5-flash-lite and 4 new xAI Grok models. Critical issue: cachedInput pricing is higher than input for grok-4-fast-reasoning and grok-4-fast-non-reasoning.

Sequence Diagram

sequenceDiagram
    participant Dev as Developer
    participant Models as models.ts
    participant App as Application
    
    Dev->>Models: Update Gemini 2.5 Pro pricing<br/>(input: 0.15→1.25, output: 0.6→10.0)
    Dev->>Models: Update Gemini 2.5 Flash pricing<br/>(input: 0.15→0.3, output: 0.6→2.5)
    Dev->>Models: Add Gemini 2.5 Flash Lite<br/>(input: 0.1, output: 0.4)
    Dev->>Models: Update Grok 4 Latest cached pricing<br/>(cachedInput: 1.5→0.75)
    Dev->>Models: Add Grok 4-0709 model
    Dev->>Models: Add Grok 4-1 Fast models<br/>(reasoning & non-reasoning)
    
    Note over Models: ⚠️ Issue: grok-4-fast-* models<br/>have cachedInput (0.25) > input (0.2)
    
    App->>Models: Request model pricing info
    Models-->>App: Return pricing (with incorrect cached values)

Loading

[greptile-apps]

Additional Comments (2)

1. apps/sim/providers/models.ts, line 839-840 (link)

   logic: cachedInput should be lower than input (0.05, not 0.25)

2. apps/sim/providers/models.ts, line 852-853 (link)

   logic: cachedInput should be lower than input (0.05, not 0.25)

_{1 file reviewed, 2 comments}

_{Edit Code Review Agent Settings | Greptile}