Skip to main content
FAQ

Which AI Models Does Prompt Anything Pro Support?

Quick Answer

Prompt Anything Pro supports 26 models across seven providers via BYOK: OpenAI, Anthropic, Google, xAI, and Groq on the cloud side, plus local models through Ollama or LM Studio and Chrome's built-in AI (Gemini Nano) on the device side. The recommended picks are OpenAI's GPT-6.1 Sol, Anthropic's Claude Opus 5.5, Google's Gemini 3.8 Flash, and xAI's Grok 4.7, alongside the most capable tiers (GPT-6 Astra, Claude Fable 5.1, Gemini 4 Argon) and faster low-cost options (GPT-6 Luna, Claude Haiku 5.5, Gemini 3.5 Flash-Lite). You configure each cloud provider's API key once — local and on-device models need no key at all — and switch models per-prompt with a one-click dropdown. The model list is fixed and updated in extension releases.

  • OpenAI (5): GPT-6.1 Sol (recommended), GPT-6 Astra (most capable), GPT-6.1 Terra (fast), GPT-6 Luna (cheapest), GPT-5.5 (previous generation).
  • Anthropic (4): Claude Opus 5.5 (recommended), Claude Sonnet 5.5 (balanced daily driver), Claude Haiku 5.5 (fast/cheap), Claude Fable 5.1 (most capable, slower).
  • Google (4): Gemini 3.8 Flash (recommended), Gemini 4 Argon (most capable), Gemini 3.6 Flash, Gemini 3.5 Flash-Lite.
  • xAI (4): Grok 4.7 (recommended), Grok 4.3, Grok 4.20 (fast, no reasoning), Grok Build 0.1 (coding). Groq (3): GPT-OSS 120B, GPT-OSS 20B, Qwen 3.8 27B (preview).
  • Local and on-device (6): Ollama or LM Studio models (Llama 3.2 3B, Llama 3.1 8B, Qwen 2.5 7B, Mistral 7B, Phi-3 Mini) and Chrome's built-in AI (Gemini Nano) — no API key, no per-token bill, nothing leaves your machine.
  • 26 models in total across 7 providers. Switch per-prompt with a single dropdown click. The list is fixed and updated in extension releases, and BYOK means you pay your provider directly.

By PlugMonkey Team, Editorial

OpenAI Models

OpenAI offers a range of models, from the most capable options to fast, cost-effective ones. Prompt Anything Pro includes five of them.
  • GPT-6.1 Sol — The recommended pick. Best all-around choice for complex reasoning, creative writing, code generation, and analysis.
  • GPT-6 Astra — The most capable OpenAI model in the list, for the hardest multi-step problems.
  • GPT-6.1 Terra — The fast tier. Great for high-volume work like summarization, rewriting, and quick Q&A.
  • GPT-6 Luna — The cheapest option, for bulk work where cost matters most.
  • GPT-5.5 — The previous generation, still selectable if you have prompts or workflows tuned to it.
  • Per-token rates differ for every model and change over time — check OpenAI's pricing page for the current numbers before you settle on a default.

Anthropic Models

Anthropic's Claude models are known for strong reasoning, nuanced writing, and reliable instruction-following. Many users prefer Claude for writing-heavy tasks and complex analysis.
  • Claude Opus 5.5 — The recommended pick. Excels at nuanced writing, long-form analysis, coding, and tasks that need careful step-by-step reasoning.
  • Claude Sonnet 5.5 — The balanced mid-tier. Close to Opus on most everyday work at a lower per-token rate, and the sensible default for daily writing and editing.
  • Claude Haiku 5.5 — Optimized for speed and cost. Ideal for quick tasks where you want Claude's writing style without paying top rates.
  • Claude Fable 5.1 — The most capable Claude model in the list, a bit slower. Worth it for the hardest problems.
  • Rates vary per model — see Anthropic's pricing page for the current per-token cost of each.

Google Models

Google's Gemini family offers competitive performance with aggressive pricing. They are especially strong at factual queries and tasks that benefit from Google's training data.
  • Gemini 3.8 Flash — The recommended pick. Fast, with strong results on long documents and everyday tasks, and the Gemini model to reach for first.
  • Gemini 4 Argon — The most capable Gemini model in the list, for the hardest problems and large datasets.
  • Gemini 3.6 Flash — A fast, lightweight tier built for high-volume simple tasks.
  • Gemini 3.5 Flash-Lite — The lightest and cheapest option, for bulk work where speed and cost come first.
  • Context limits and per-token rates differ across the family — Google's pricing page has the current figures.

xAI Models

xAI's Grok family is the newest addition to the cloud line-up. It is worth reaching for when you want a different reasoning style from the OpenAI/Anthropic/Google trio, or a second opinion on a hard problem.
  • Grok 4.7 — The recommended pick, and the Grok model to start with.
  • Grok 4.3, Grok 4.20 (fast, no reasoning), and Grok Build 0.1 (coding) are selectable from the same dropdown once your xAI key is configured.
  • Per-token rates are published on xAI's API page — check there before setting Grok as a default.

Groq

Groq is an inference provider rather than a model lab: it serves open-weight models on custom hardware at very high token throughput. Use it when latency matters more than frontier reasoning quality — bulk summarizing, translation passes, or anything you run dozens of times an hour.
  • Three fast open-weight models served at a fraction of flagship latency: GPT-OSS 120B, GPT-OSS 20B, and Qwen 3.8 27B (preview).
  • Same BYOK setup as every other cloud provider: paste a key once, then pick the model per prompt.
  • Current model list and rates live in Groq's model documentation.

Local and On-Device Models (No Key Required)

Two of the seven providers never touch the network. If you run a model locally through Ollama or LM Studio, or use Chrome's built-in AI, the text you select never leaves your machine — there is no API key, no per-token bill, and no external privacy policy to trust. The trade-off is capability: a local model will not match GPT-6.1 Sol or Claude Opus 5.5 on hard reasoning, so the practical pattern is on-device for anything sensitive and a cloud key for anything that needs frontier quality.
  • Ollama — Point Prompt Anything Pro at your local Ollama server. The list includes Llama 3.2 3B, Llama 3.1 8B, Qwen 2.5 7B, Mistral 7B, and Phi-3 Mini.
  • LM Studio — Same idea with LM Studio's local server, if that is your preferred runner.
  • Chrome's built-in AI (Gemini Nano) — Runs inside the browser itself. Nothing to install, no key, no network round-trip.
  • Zero marginal cost: local and on-device prompts do not appear on any provider invoice.

How to Switch Between Models

Prompt Anything Pro makes model switching effortless. In the extension settings, you configure API keys for each provider you want to use. When you trigger a prompt (by highlighting text and clicking the floating button, or pressing ⌘J / Alt+J in a text field), you can select which model to use from a dropdown. You can also set a default model that is used automatically unless you override it. Switching models takes one click — there is no need to change settings pages or reconfigure anything.
Prompt Anything Pro's built-in template library with Summarize, Explain, Writing, and Translate categories

Which Model Should You Use?

Different models have different strengths. Here is a practical guide for choosing the right model for common tasks.
  • Complex analysis or reasoning — GPT-6.1 Sol or Claude Opus 5.5. Both are top-tier; try both and see which style you prefer. For the hardest multi-step problems, GPT-6 Astra is worth a look too.
  • Creative writing and editing — Claude Opus 5.5 and Claude Sonnet 5.5 tend to produce more natural, nuanced prose. GPT-6.1 Sol is also excellent but can read as more formulaic.
  • Code generation and debugging — GPT-6.1 Sol and Claude Opus 5.5 are both strong, and Grok Build 0.1 is tuned for coding. Run the same prompt through each on a task you already know the answer to, then keep whichever you trust more.
  • Summarization and quick tasks — GPT-6.1 Terra, Claude Haiku 5.5, or Gemini 3.8 Flash. Fast and cheap, more than capable for straightforward work.
  • Long document processing — Gemini 3.8 Flash or Gemini 4 Argon. Context limits differ per model, so check the provider's model card before pasting something very large.
  • Budget-conscious daily use — GPT-6 Luna or Gemini 3.5 Flash-Lite for most tasks, escalating to GPT-6.1 Sol or Claude Opus 5.5 only for complex work.

How New Models Are Added

The AI landscape moves fast, with new models launching regularly. Prompt Anything Pro offers a fixed list of 26 models, and PlugMonkey updates that list in extension releases (the latest is version 1.9.10). A model a provider has just launched appears in the extension only after we add it. Because the extension uses the BYOK (Bring Your Own Key) approach, you still pay your provider's API rates directly.

Want a Second Opinion?

Ask AI for an independent perspective on this question.

AI responses are generated independently and may vary

Try Prompt Anything Pro Free

Access GPT-6.1 Sol, Claude Opus 5.5, Gemini 3.8 Flash, and Grok 4.7 from any webpage. Bring your own API keys, pay actual costs.

5.0/5 (3 ratings)20 active users