SI Briefing
September 2026's Model Rush: GPT-6, Claude 5.5, Gemini 4 Argon and Grok 4.7
On this page
If it felt like a new frontier model launched every few days in September 2026, that’s because one did. Here’s the month in order.
The releases
| Date | Lab | Model | Price (in / out per 1M) |
|---|---|---|---|
| Sep 1 | Anthropic | Claude Fable 5.1 | $10 / $50 |
| Sep 2 | Gemini 3.8 Flash | $0.75 / $3.75 (intro) | |
| Sep 2 | Meta | Muse Spark 1.3 | $1.25 / $4.25 |
| Sep 3 | OpenAI | GPT-6 Astra | $10 / $50 |
| Sep 21 | xAI | Grok 4.7 | $2 / $6 (under 200K) |
| Sep 22 | Anthropic | Claude Opus 5.5 | $4 / $20 |
| Sep 22 | OpenAI | GPT-6 Sol and GPT-6 Luna | Luna: $0.10 / $0.50 |
| Sep 28 | Anthropic | Claude Sonnet 5.5 | $2 / $10 |
| Sep 29 | OpenAI | GPT-6.1 Sol | $2 / $10 |
| Sep 30 | Gemini 4 Argon (limited preview) | Not yet public |
Three takeaways
- The flagship price point settled at $10/$50. Both GPT-6 Astra and Claude Fable 5.1 list at the same price, while near-flagship models (Opus 5.5, GPT-6.1 Sol, Sonnet 5.5) cost 60–80% less.
- Million-token context is now standard. Every OpenAI, Anthropic and generally available Google model on this list handles about a million tokens (Google has not published Gemini 4 Argon’s input context). Grok 4.7 offers 500K.
- Cyber capability is shaping rollouts. GPT-6 Astra and Gemini 4 Argon both went first to vetted cyber-defenders before wider release. OpenAI also cancelled a planned GPT-6.1 Astra launch after safety tests. That caution sits in the background of the White House Accord on Super Intelligence, signed on September 29.
Compare them
See the full model tracker and head-to-heads like GPT-6 Astra vs Claude Fable 5.1 and GPT-6.1 Sol vs Claude Sonnet 5.5.
Sources
- Claude models overview — Anthropic
- OpenAI API models — OpenAI
- xAI models — xAI
- Gemini API pricing — Google
Written by
Luka Kušec · Editor
Editor of SI.info. Writes about Super Intelligence, technology policy and the people building frontier models.