GPT-6 Astra and 3 More Models Just Landed in AI Optimise
By Emmanuel Abou Chabke, Founder and Editor · Reviewed by the AIQuickPrompt editorial team · About the publisher
AI Optimise has four new brains. GPT-6 Astra joins the premium tier as the most capable model we offer, the latest ChatGPT Instant model arrives for natural conversational rewrites, and Gemini 3.8 Flash and Gemini 3.7 Flash join the free fast tier that every plan can use. That takes AIQuickPrompt from 39 to 43 selectable models. This short guide explains what changed, what each new model is good at, and how to pick without overthinking it.
What is new in AI Optimise
Every model in AIQuickPrompt is available from the same place: open a prompt card, press AI Optimise, and choose a model from the dropdown. Nothing about that flow changed. The list simply got better at the top and at the fast end.
The four additions are live now for every account, with the usual rules: fast models are free on every plan, premium models use a Pro credit or a top-up credit, and bring-your-own-key models stay free of our credits entirely.
- GPT-6 Astra | premium. The newest and most capable ChatGPT model for complex reasoning and demanding rewrites.
- ChatGPT Latest Instant | premium. The conversational ChatGPT model, tuned for natural, human-sounding phrasing.
- Gemini 3.8 Flash | fast and free. The newest Gemini Flash generation, quick and surprisingly strong.
- Gemini 3.7 Flash | fast and free. A fast, smart alternative when you want a second opinion at no cost.
When should you reach for GPT-6 Astra?
Astra is the model to use when the prompt itself is the hard part: long system prompts, multi-step agent instructions, prompts that must hold several constraints at once, or a brief where a subtle misreading costs you real work later. It reasons before it answers, so it catches contradictions and vague instructions that faster models quietly paper over.
It is a premium model, so each run costs one credit. That is the right trade for a prompt you will reuse a hundred times, and the wrong trade for tidying a two-line request. Use a fast model for the small stuff and save Astra for the prompts that earn their keep.
- Long or layered system prompts with several rules to respect at once.
- Prompts that feed an automation, an agent or a team workflow.
- High-stakes client-facing briefs where tone and precision both matter.
Which tier does each new model sit in?
Tiers decide what a run costs you, not how good a model is. Fast models are free on every plan, with three runs a day on Free and fair-use unlimited on Pro. Premium models draw on your Pro credits or a €5 top-up of 25 credits that never expire. Bring-your-own-key models run on your own OpenAI, Anthropic or Google AI Studio key and never touch our credits at all.
In practice most people run the fast tier all day and spend a premium credit a handful of times a week, which is exactly how the pricing is designed to work.
- Free every day: Gemini 3.8 Flash, Gemini 3.7 Flash, Gemini 3.6 Flash, GPT-5.6 Luna, GPT-5.4 Mini and the rest of the fast tier.
- One credit per run: GPT-6 Astra, ChatGPT Latest Instant, GPT-5.6 Sol, GPT-5.5 Pro, Gemini 3.1 Pro and the other premium models.
- No credits at all: Claude, OpenAI and Gemini models on your own API key, billed by your provider.
A five-second way to choose
Ask one question: does this prompt need judgement, or just tidying? Tidying means a fast model. Judgement means a premium one. That single rule covers almost every case and stops you burning credits on work a free model handles perfectly.
If you optimise prompts all day, connect your own API key in Settings. You keep every model, remove every daily limit, and pay your provider directly at cost.