The state of evaluation in 2026
The state of evaluation in 2026. A short analysis built from the index, with the model-level data behind the claim.
Where things stand
Evaluation is no longer a single market. The index now tracks 600 models from 40 providers, and the spread between the cheapest and most expensive option is wider than most teams expect.
For evaluation, the practical question is not 'which model is best' but 'which model is best at a price I can defend'. The data below shows how the leading candidates compare.
What the data shows
Among the models most relevant to evaluation, GPT-5.2 Pro has the strongest average score at 89.6, while GPT-5.1 Reasoning is the cheapest at $20.27 per million input tokens. The gap between them is the trade-off most teams are actually managing.
Prices move faster than benchmark scores. A model that looked expensive six months ago may now be mid-market, and a new release can reset the value calculation overnight.
What to expect next
- More price competition in the mid-tier.
- Longer context windows at lower prices.
- More open-weight models that are good enough for production.
- Benchmarks that focus on agents and long-horizon tasks.
Models relevant to evaluation
| Model | Provider | Input / 1M | Context | Avg score |
|---|---|---|---|---|
| GPT-5.2 Pro | OpenAI | $21.32 | 2,000,000 | 89.6 |
| GPT-5.2 Max | OpenAI | $21.39 | 1,000,000 | 89.2 |
| Claude Opus 4.5 Pro | Anthropic | $20.98 | 200,000 | 89.1 |
| GPT-5.1 Reasoning | OpenAI | $20.27 | 400,000 | 88.3 |
| Claude Opus 4.5 Reasoning | Anthropic | $20.91 | 2,000,000 | 88.2 |
Related
FAQ
What is the main takeaway of this evaluation analysis?
Evaluation is no longer a single market. The index now tracks 600 models from 40 providers, and the spread between the cheapest and most expensive option is wider than most teams expect.
Are the figures official?
No. The figures come from the AI & LLM Index demo dataset. Prices and scores are compiled for comparison; benchmark values are modelled estimates.
How should I use this in a buying decision?
Use it to build a shortlist, then test the shortlist on your own examples. Check the model pages for current price and context details.