The models we use
The right model for the job.
New AI models ship every month. We track them, test them, and put the right one behind each part of your business, so you get the outcome without keeping up with it yourself.
Updated 21 November 2026
By job
Which model does which job.
Running cost
Cost to answer 1,000 customer enquiries.
Estimate In US dollars, at each company’s published standard API price.
| Model | Company | Estimated cost (USD) |
|---|---|---|
| GPT-6 Luna | OpenAI | $0.55 |
| Gemini 3.8 Flash | $4.13 | |
| GPT-6.1 Terra | OpenAI | $5.50 |
| Gemini 4 Hydrogen | $8.25 | |
| Claude Sonnet 5.5 | Anthropic | $11 |
| GPT-6.1 Sol | OpenAI | $11 |
| Gemini 4 Carbon | $11 | |
| Gemini 4 Argon | $11 | |
| Gemini 4 Neon | $17 | |
| Claude Opus 5.5 | Anthropic | $22 |
| Claude Fable 5.2 | Anthropic | $44 |
| GPT-6 Astra | OpenAI | $55 |
What’s one extra job worth?
Running cost uses GPT-6.1 Sol, the middle of the chart above, at its official price checked 21 November 2026, and the same enquiry size as the chart. It’s an estimate of the AI’s cost only.
The AI itself costs little to run. Most of the cost is building and running the system around it. We quote that after a 60-minute consultation.
Talk to AETHERCurrent models
The models we work with now.
Checked every day against each company’s own announcements, documentation and changelogs. New generally available models are added automatically from those sources only.
OpenAI
- GPT-6 AstraReleased 3 September 2026
The most demanding work: complex reasoning, research, computer use and document creation.
Official page for GPT-6 Astra (opens in a new tab) ↗ - GPT-6.1 SolReleased 29 September 2026
Complex coding and professional work, at a lower cost than GPT-6 Astra.
Official page for GPT-6.1 Sol (opens in a new tab) ↗ - GPT-6 LunaReleased 22 September 2026
Focused, high-volume tasks where speed and running cost matter most.
Official page for GPT-6 Luna (opens in a new tab) ↗ - GPT-Live 1Released 10 September 2026
Natural voice conversations that handle interruptions, for phone and voice agents.
Official page for GPT-Live 1 (opens in a new tab) ↗ - GPT-TranscribeReleased 28 July 2026
Accurate speech-to-text for calls, meetings and recordings.
Official page for GPT-Transcribe (opens in a new tab) ↗ - GPT-6.1 TerraReleased 26 October 2026
Handy for reading photos and drawings, working through on-screen apps for you, and writing or editing text quickly and affordably.
Official page for GPT-6.1 Terra (opens in a new tab) ↗
Anthropic (Claude)
- Claude Opus 5.5Released 22 September 2026
Long, sprawling jobs: large sets of documents, audits and changes across a whole codebase.
Official page for Claude Opus 5.5 (opens in a new tab) ↗ - Claude Sonnet 5.5Released 28 September 2026
Everyday business work: fixing bugs and drafting polished documents, slides and spreadsheets.
Official page for Claude Sonnet 5.5 (opens in a new tab) ↗ - Claude Haiku 5.5Released 7 October 2026
Quick, low-cost replies for busy, repetitive work: summarising, sorting requests and live customer chat.
Official page for Claude Haiku 5.5 (opens in a new tab) ↗ - Claude Fable 5.2Released 9 November 2026
Suited to demanding software projects and financial analysis, including long jobs it can keep working on by itself.
Official page for Claude Fable 5.2 (opens in a new tab) ↗
Google (Gemini)
- Gemini 3.8 FlashReleased 2 September 2026
Agents and complex multi-step workflows at Flash speed.
Official page for Gemini 3.8 Flash (opens in a new tab) ↗ - Gemini 3.8 LiveReleased 15 September 2026
Low-latency voice agents for real-time conversations.
Official page for Gemini 3.8 Live (opens in a new tab) ↗ - Gemini 3.5 TranscribeReleased 26 August 2026
Speech-to-text with speaker labels and timestamps, for call and meeting notes.
Official page for Gemini 3.5 Transcribe (opens in a new tab) ↗ - Gemini 4 HydrogenReleased 12 November 2026
Fast, low-cost help for everyday coding and agent workflows, with room for very long documents.
Official page for Gemini 4 Hydrogen (opens in a new tab) ↗ - Gemini 4 CarbonReleased 10 November 2026
Google's most capable model, for hard software, maths and science problems and long multi-step agent work.
Official page for Gemini 4 Carbon (opens in a new tab) ↗ - Gemini 4 ArgonReleased 13 October 2026
Reasoning with quicker answers and lower running cost than Carbon, for everyday agents and analysis.
Official page for Gemini 4 Argon (opens in a new tab) ↗ - Gemini 4 NeonReleased 18 November 2026
Suits small businesses needing complicated coding and knowledge work finished quickly from start to finish.
Official page for Gemini 4 Neon (opens in a new tab) ↗
Beyond the model
What we connect the model to.
We connect it to your CRM, email, calendar, database and tools. We give it skills, which are repeatable procedures, and your business rules. And we keep a person in the loop for anything that matters.
When a better model ships, we test it on your real work and switch if it’s better, so you’re never locked into one company’s model.
Start here
What should be working better?
Model names are trademarks of their owners. AETHER is independent and not affiliated with or endorsed by these companies. Last reviewed: 21 November 2026.