Best value AI models: 43 models priced against performance
The measured performance score set against the price per million tokens, input and output weighed 3 to 1: the models no cheaper model outscores, first at list price, then at the price you pay through AIGROW API credit.
Run Claude Fable 5.2 up to 60% below list price. $25 a month brings $110 of API credit and cancels any time; $25 once buys $100. Scores from the Epoch Capabilities Index by Epoch AI, data of 12 November 2026, under CC BY 4.0. Prices from each maker’s pricing page, read on 12 November 2026.
- Best value at list price
- 4
- Best value with the credit
- 6
- Models priced
- 43
- Data of
- 12 November 2026
Input and output weighed 3 to 1.
At list price, 4 models are the best value on the Epoch Capabilities Index (12 November 2026): for each of them, no cheaper model scores higher. They run from Claude Haiku 5.5 at $0.20 to Claude Fable 5.2 at $16 per million tokens, input and output weighed 3 to 1. Through AIGROW API credit, Claude Fable 5.2 costs $6.40.
At list price
The models no cheaper model outscores on the index, cheapest first, at the price each maker lists. Every other model costs more than one of these and scores no higher.
| Model | Index score | Rank | List price, per 1M tokens, 3:1 |
|---|---|---|---|
| Claude Haiku 5.5Anthropic | 160.4 | 18 of 256 | $0.20 |
| GPT-6.1 TerraOpenAIRun GPT-6.1 Terra up to 75% below list priceCertified genuine by AIGROW on 11 November 2026 | 167.7 | 6 of 256 | $2.00 |
| Gemini 4 CarbonGoogle DeepMind | 169.8 | 2 of 256 | $4.00 |
| Claude Fable 5.2AnthropicRun Claude Fable 5.2 up to 60% below list priceCertified genuine by AIGROW on 11 November 2026 | 173.4 | 1 of 256 | $16.00 |
With AIGROW API credit
The same rule at the price you pay: through AIGROW API credit for the models it sells, at list for the rest. A cheaper way to the same score moves the line.
| Model | Index score | Rank | Price paid, per 1M tokens, 3:1 |
|---|---|---|---|
| Claude Haiku 5.5Anthropic | 160.4 | 18 of 256 | $0.20 |
| GPT-6.1 TerraOpenAIRun GPT-6.1 Terra up to 75% below list priceCertified genuine by AIGROW on 11 November 2026 | 167.7 | 6 of 256 | |
| GPT-6.1 SolOpenAIRun GPT-6.1 Sol up to 75% below list priceCertified genuine by AIGROW on 11 November 2026 | 168.5 | 5 of 256 | |
| Claude Opus 5.5AnthropicRun Claude Opus 5.5 up to 60% below list priceCertified genuine by AIGROW on 11 November 2026 | 168.5 | 4 of 256 | |
| Gemini 4 CarbonGoogle DeepMind | 169.8 | 2 of 256 | $4.00 |
| Claude Fable 5.2AnthropicRun Claude Fable 5.2 up to 60% below list priceCertified genuine by AIGROW on 11 November 2026 | 173.4 | 1 of 256 |
Certified genuine by AIGROW means the model passed AIGROW’s own check against the same model at its maker on 11 November 2026. No model maker reviewed or endorsed it.
How to read this ranking
What the scores measure, what they leave out, and where each number comes from.
- What it measures
- The measured performance score set against the price per million tokens, input and output weighed 3 to 1: the models no cheaper model outscores, first at list price, then at the price you pay through AIGROW API credit.
- What it leaves out
- A price per token is not a price per task: a model that writes longer answers or thinks longer costs more for the same question.
- The scale
- Epoch AI sets the scale so that Claude 3.5 Sonnet scores 130 and GPT-5 scores 150, and it has no maximum. A score means something only next to another, and a model scores higher by doing better on harder tests.
- How to read the margin
- Each score carries a 90% interval: under Epoch AI’s model, the true score lies inside it nine times in ten. Where two models’ intervals overlap, the data cannot say which is better. Rows are still numbered by score, and a row says how many models above it overlap.
- How the index is built
- Epoch AI scores a model only once it has results on at least 4 benchmarks, keeps its best result on each across reasoning settings, and refits every score when new results arrive, so a score can move a little even when a model’s own results have not changed. The method: Epoch Capabilities Index.
- Not updated in over 120 days
- A result run more than 120 days before the newest data on these pages carries this mark, so an older result can be weighed as one.
- Prices
- The list price per million tokens from the maker’s own pricing page, read on 12 November 2026. It is a price per token, not per answer: a higher reasoning setting spends more tokens on the same question. Where AIGROW API credit sells a model, the list price is struck through and the price through the credit sits under it.
Source: Epoch AI, benchmark data (data of 12 November 2026), licensed under CC BY 4.0. We change no score: we add the prices and the links to our model pages.
What each of them costs to run
The price list works three everyday jobs on every model, and our own benchmark runs them on real business tasks to see which pass.
Data of 12 November 2026