Hi @vagrant egret. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
#feedback: Warp Agentic Coding comparing to my Claude Code experience
bleak orbitBOT
Thank you for your detailed feedback and for sharing your experience comparing Warp’s Agentic Coding to Claude Code and Codex.
Opus 4 Model Support:
Currently, Warp supports a curated set of LLMs for Agent Mode, including Claude 3.7 Sonnet, Claude 3.5 Sonnet, Claude 3.5 Haiku, OpenAI GPT-4o, GPT-4.1, Gemini 2.5 Pro, and others. As of the latest documentation, Claude Opus 4 is not listed as an available model in Warp’s engine list. Warp does use Sonnet 4 (Claude 3.7 Sonnet) as the default and allows you to select from the supported models via a dropdown in Agent Mode. There’s no mention of quota-based automatic fallback between models (like Opus to Sonnet) or Opus 4 availability, but your suggestion aligns with recent model experimentation by the Warp team, who have evaluated Opus 4 internally. However, they observed that Sonnet 4 performed better for their use cases, and Opus 4 is not yet available for user selection. If you’d like to see Opus 4 added, the best path is to submit this as a feature request via their GitHub issues board or feedback channels Model Choice, Terminal-Bench.
Diff Navigation with Keyboard:
You are not missing anything—Warp’s built-in diff editor allows you to navigate between diff hunks using the UP and DOWN arrow keys, and between files with LEFT and RIGHT. You do not need to switch into editor mode just to scroll through diffs; editor mode is only needed for manual code edits. If you’re experiencing issues with navigation, it may be a bug or a UI limitation worth reporting Reviewing Agent-generated Code.
Diff Color Contrast:
Your feedback about the red, yellow, and green colors in the diff view being hard to distinguish is valid. The knowledge sources do not mention any current settings to adjust diff color schemes or contrast, but this is valuable feedback for the team. You can submit UI/UX suggestions via their GitHub or feedback form.
Python "->" Token Bug:
There is no explicit mention of this bug in the knowledge sources. If you’re seeing unprintable characters replacing the "->" token in generated Python code, it’s best to report this as a bug with a screenshot to help the team reproduce and fix it.
Multiple Profiles/Accounts:
Warp currently does not support running multiple profiles or being logged into two accounts simultaneously (e.g., personal and work) in separate windows. The knowledge sources do not mention this as a feature, but your use case is a common request and worth submitting as a feature suggestion FAQ.
Pro Plan Trial:
There is no mention in the knowledge sources of a free trial for the Pro plan. Warp’s Free plan includes 150 AI requests per month, and you can upgrade to Pro or Turbo for higher limits. Pay-as-you-go overages are available after you hit your quota. For trial or evaluation requests, you may want to contact Warp directly via their support channels or contact sales Pricing.
General Experience:
It’s great to hear that you find Warp’s integration and workflow superior to Claude and Codex. Warp’s agentic coding platform is designed for deep terminal integration, multi-agent management, and a modern, prompt-driven development experience Warp 2.0: Reimagining coding.
If you’d like to provide more feedback or track feature requests, the GitHub issues board is the recommended place.