This page tracks the models that matter right now, across OpenAI, Anthropic, Google and the open-weight leaders. It is updated when a vendor ships something new, so the table stays current instead of turning into another stale “new AI models” post.
Prices are list prices per million tokens, checked on 7 November 2026. Buying through a cloud marketplace can cost more than buying direct, and batch tiers are usually half price.
For the head-to-head view of the three default flagships, see the AI model comparison for 2026. For what the benchmark numbers mean, see AI model benchmarks explained.
The models most teams are choosing between
| Model | Vendor | Listed | Context | Price (input / output) | Direct price |
|---|---|---|---|---|---|
| GPT-6.1 Sol | OpenAI | 29 Sep 2026 | 1.05M | $2 / $10 | $1 / $5 |
| GPT-6.1 Terra | OpenAI | 26 Oct 2026 | 1.05M | $1 / $5 | $0.50 / $2.50 |
| Claude Fable 5.1 | Anthropic | 1 Sep 2026 | 1M | $10 / $50 | $10 / $50 |
| Gemini 3.6 Flash | 21 Jul 2026 | 1.05M | $0.75 / $3.75 | $0.75 / $3.75 |
The “listed” price is what marketplaces such as OpenRouter show. OpenAI’s direct API is cheaper than the marketplace rate for both GPT-6.1 models, so check both before you budget.
OpenAI
| Model | Listed | Context | Price | Notes |
|---|---|---|---|---|
| GPT-6.1 Sol | 29 Sep 2026 | 1.05M | $2 / $10 | Flagship; $1 / $5 direct |
| GPT-6.1 Terra | 26 Oct 2026 | 1.05M | $1 / $5 | Cheaper tier; $0.50 / $2.50 direct; computer use |
| GPT-6 Astra | 4 Sep 2026 | 1.05M | $10 / $50 | Premium tier, same price band as Fable |
| GPT-6 Luna | 22 Sep 2026 | 1.05M | $0.10 / $0.50 | High-volume tier |
| GPT-6 Sol | 22 Sep 2026 | 1.05M | $2 / $10 | Superseded by 6.1 Sol |
| GPT-5.6 Sol | June 2026 | 1.1M | $5 / $30 | Previous generation |
| GPT-5.6 Terra | June 2026 | 1.1M | $2.50 / $15 | Previous generation |
| GPT-5.6 Luna | June 2026 | 1.1M | $1 / $6 | Previous generation |
The pattern is clear: OpenAI ships a new tier roughly monthly, and each generation lands cheaper per token than the one it replaces. Moving from GPT-5.6 Sol to GPT-6.1 Sol cuts the direct price by about 80%.
Read the GPT-6.1 Sol vs Terra breakdown before choosing between the two current tiers.
Anthropic
| Model | Listed | Context | Price | Notes |
|---|---|---|---|---|
| Claude Fable 5.1 | 1 Sep 2026 | 1M | $10 / $50 | Flagship; cache reads $0.25 |
| Claude Mythos 5.1 | Sep 2026 | 1M | Restricted | Same model, fewer safeguards; vetted access |
| Claude Opus 5.5 | 22 Sep 2026 | 1M | $4 / $20 | Workhorse flagship |
| Claude Sonnet 5.5 | 28 Sep 2026 | 1M | $2 / $10 | Balanced tier |
| Claude Haiku 5.5 | 7 Oct 2026 | 1M | $0.10 / $0.50 | Cheapest Claude tier |
| Claude Opus 5 | 24 Jul 2026 | 1M | $5 / $25 | Previous Opus |
| Claude Fable 5 | 9 Jun 2026 | 1M | $10 / $50 | Previous Fable |
| Claude Opus 4.8 | 27 May 2026 | 1M | $5 / $25 | Previous Opus |
| Claude Sonnet 4.6 | June 2026 | 1M | $3 / $15 | Previous Sonnet |
| Claude Haiku 4.5 | Oct 2025 | 200K | $1 / $5 | Previous Haiku |
Anthropic’s September and October releases were mostly about cost, not headline prices. Fable 5.1 kept $10 / $50 but cut cache reads to $0.25 per million tokens. The Fable 5.1 upgrade notes explain who benefits.
See the full Claude model lineup for positioning.
| Model | Listed | Context | Price | Notes |
|---|---|---|---|---|
| Gemini 4 Argon | 13 Oct 2026 | 1.05M | $2 / $10 | Newest listed Gemini |
| Gemini 3.6 Flash | 21 Jul 2026 | 1.05M | $0.75 / $3.75 | Volume and multimodal |
| Gemini 3.5 Flash | 19 May 2026 | 1.05M | $1.50 / $9.00 | Previous Flash |
| Gemini 3.5 Flash-Lite | 21 Jul 2026 | 1.05M | $0.30 / $2.50 | Cheapest tier |
Gemini 3.6 Flash halved the launch price of 3.5 Flash. The Gemini 3.6 Flash guide covers batch pricing and where it fits.
The Gemini lineup page tracks the rest of the family.
Open-weight and budget models
| Model | Vendor | Listed | Context | Price |
|---|---|---|---|---|
| Nemotron 3 Ultra | NVIDIA | 4 Jun 2026 | 262K | $0.60 / $2.40 |
| Nemotron 3.5 Super VL | NVIDIA | 15 Oct 2026 | 1M | $0.15 / $0.50 |
| GLM 5.3 | Z.ai | 18 Aug 2026 | 1M | $0.07 / $1.40 |
| GLM 5.3 Flash | Z.ai | 26 Aug 2026 | 1M | $0.15 / $0.50 |
| GLM 5.3 Prime | Z.ai | 23 Sep 2026 | 1M | $2.80 / $8.80 |
Open-weight models are worth pricing even if you use a hosted API. They set the floor for what a capable model should cost.
What changed in the last month
- 26 October 2026: OpenAI listed GPT-6.1 Terra, a cheaper tier with computer use.
- 13 October 2026: Google listed Gemini 4 Argon.
- 7 October 2026: Anthropic listed Claude Haiku 5.5 at $0.10 / $0.50.
- 1 September 2026: Anthropic shipped Claude Fable 5.1 with cheaper cache reads.
- September 2026: Anthropic introduced Claude Mythos 5.1 for vetted cyber and life-science teams.
How to read this page
- Dates are when a model appeared in the OpenRouter catalogue. Vendor announcements can be a few days earlier.
- Prices are list prices per million tokens. Cached, batch and priority tiers differ.
- Context is the maximum input window, not the useful window. A 1M window does not mean the model stays accurate across all of it.
- Restricted means the model is not generally available through public APIs.
FAQ
What is the newest AI model right now?
As of 7 November 2026, the newest listed releases are GPT-6.1 Terra (26 October), Gemini 4 Argon (13 October) and Claude Haiku 5.5 (7 October).
Which is the cheapest capable model?
Among the major vendors, Claude Haiku 5.5 and GPT-6 Luna are in the same $0.10 / $0.50 band. Gemini 3.6 Flash is the cheapest of the flagship-class models at $0.75 / $3.75.
Why do OpenAI prices differ between sources?
OpenAI’s direct API is cheaper than marketplace listings for GPT-6.1 Sol and Terra. The listed price includes the marketplace margin.
Is this page updated automatically?
No. It is checked and updated by hand when a vendor ships a model, so every price has a date next to it.
Next step
If you are choosing a model for a product, the AI model comparison and the benchmark guide cover the trade-offs. To test a shortlist on your own data, contact 24Bit System.
