Pains

When users change passwords, refresh tokens may be revoked by OAuth providers, forcing applications to handle re-authentication. Graceful re-auth flows must be implemented to prevent user frustration.

Gemini CLI does not automatically save sessions, requiring users to manually manage conversation state. This reduces continuity and forces workarounds to maintain context across CLI invocations.

Gemini models (especially 2.5-Pro) generate code with excessive comments, which is good for learning but poor for production agentic coding. The verbose commenting style makes AI-generated code immediately identifiable and impractical for autonomous agents.

Safety settings in the Gemini API have unexpected behavior and are inadequately documented compared to other vendors, causing confusion during implementation.

Google offers Gemini API access through two confusing product names (AI Studio and Vertex AI) with unclear differences for developers. This naming confusion adds unnecessary cognitive load during onboarding and increases time to first API call.

Developers lack awareness about available npm security tools, with some respondents admitting they don't know what options exist. This contributes to only 40% satisfaction with current security tools despite available solutions.

Gemini API requires internet connectivity with no local or offline alternative. Applications cannot function without a network connection.

Phone support is only available for enterprise contracts, leaving smaller teams and individual developers without direct communication channels for critical issues. This limits support options compared to competitors offering broader support tiers.

Python lacks built-in linear algebra functionality, requiring developers to rely on external libraries like NumPy for mathematical operations.

TypeScript requires explicit type annotations and generates boilerplate code for common patterns (interfaces, generics), resulting in significantly more verbose code compared to JavaScript.

TypeScript tooling never suggests how to type something properly, unlike other typed languages such as ReasonML and Flow. Developers must manually hunt down and write correct type definitions, which is time-consuming and error-prone.

Developers report needing better testing solutions and CI/CD tools, but these aren't always flagged as primary blockers. Unclear signal about where tool investment matters most.

TypeScript adds a compilation step to the development workflow that can take seconds or minutes in large projects, slowing down the development cycle. While modern tools like esbuild, swc, and Vite help reduce build times, this overhead remains a consideration for project setup.

Claude 3.5 Sonnet averages 3-5 seconds for complex responses versus GPT-4's 2-3 seconds. This latency difference becomes noticeable in real-time chat applications and high-frequency API calls, impacting user experience in customer service scenarios.

Anthropic's safety-first approach results in overly cautious responses for creative writing, marketing content, and edgy humor. Users report 23% more declined requests compared to GPT-4 for legitimate creative tasks, frustrating marketing and creative professionals.

Some OAuth providers (e.g., Box) require HTTPS for redirect URIs, making it difficult or impossible to test locally with HTTP URLs like `http://localhost:5001`.

Anthropic's prompt caching has a 5-minute time-to-live, meaning low-traffic endpoints may not see consistent cache hits. Even minor whitespace changes invalidate cached prefixes, requiring exact matching across calls.

TypeScript reports compilation errors for hardcoded values that clearly match their expected types (e.g., a hardcoded string 'compact' matching the literal type 'compact'), causing false positive errors and frustration.

Poor redirect flow performance or UX causes user frustration and abandonment. Developers must optimize and clarify the post-authorization redirect experience.

Users struggle to revoke application access if the revocation flow is not straightforward. Poor revocation UX reduces user control over permissions.

Developers must implement robust error handling covering multiple HTTP status codes (400, 403, 429, 500) with different retry strategies. Implementing exponential backoff and graceful error catching adds complexity to error handling logic.

TypeScript's standard library type definitions have not kept pace with JavaScript's TC39 proposals. Temporal began shipping in JavaScript engines but TypeScript issue #60164 shows the type definitions are still missing, creating a gap between runtime capabilities and type safety.

OpenAI API unexpectedly auto-generates questions and answers during calls, producing output that wasn't requested and requiring developers to implement additional filtering logic.

GPT-5 can feel slower than GPT-4o for simpler, everyday queries and coding tasks. Community backlash occurred regarding performance degradation for simple coding tasks before OpenAI fine-tuned model routing.