Compare two Claude models
Fable 5.1vsOpus 5.5
Opus 5.5 at Medium gives the same quality as Fable 5.1 at Max for 8.3× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5.1 | Opus 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 70× | 64.8 % | 25× | 66.0 % |
| medium | 84× | 67.2 % | 35× | 70.8 % |
| high | 102× | 68.2 % | 43× | 72.2 % |
| xHigh | 186× | 69.8 % | 74× | 73.9 % |
| max | 293× | 70.6 % | 194× | 74.7 % |
Select two models to compare.
Fable 5.1vsFable 5
Fable 5.1 at Medium gives the same quality as Fable 5 at Max for 3.2× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5.1 | Fable 5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 70× | 64.8 % | 69× | 60.5 % |
| medium | 84× | 67.2 % | 95× | 64.0 % |
| high | 102× | 68.2 % | 115× | 65.9 % |
| xHigh | 186× | 69.8 % | 165× | 66.9 % |
| max | 293× | 70.6 % | 270× | 67.8 % |
Fable 5.1vsOpus 5.5
Opus 5.5 at Medium gives the same quality as Fable 5.1 at Max for 8.3× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5.1 | Opus 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 70× | 64.8 % | 25× | 66.0 % |
| medium | 84× | 67.2 % | 35× | 70.8 % |
| high | 102× | 68.2 % | 43× | 72.2 % |
| xHigh | 186× | 69.8 % | 74× | 73.9 % |
| max | 293× | 70.6 % | 194× | 74.7 % |
Fable 5.1vsOpus 5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5.1 | Opus 5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 70× | 64.8 % | 39× | 62.5 % |
| medium | 84× | 67.2 % | 57× | 66.1 % |
| high | 102× | 68.2 % | 84× | 68.0 % |
| xHigh | 186× | 69.8 % | 111× | 69.2 % |
| max | 293× | 70.6 % | 148× | 69.6 % |
Fable 5.1vsSonnet 5.5
Sonnet 5.5 at xHigh gives the same quality as Fable 5.1 at Max for 8.9× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5.1 | Sonnet 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 70× | 64.8 % | 12× | 58.5 % |
| medium | 84× | 67.2 % | 13× | 61.7 % |
| high | 102× | 68.2 % | 18× | 66.9 % |
| xHigh | 186× | 69.8 % | 33× | 70.9 % |
| max | 293× | 70.6 % | 115× | 72.8 % |
Fable 5.1vsSonnet 5
Fable 5.1 at Low gives the same quality as Sonnet 5 at Max for 1.6× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5.1 | Sonnet 5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 70× | 64.8 % | 16× | 48.3 % |
| medium | 84× | 67.2 % | 23× | 52.8 % |
| high | 102× | 68.2 % | 33× | 55.9 % |
| xHigh | 186× | 69.8 % | 50× | 57.5 % |
| max | 293× | 70.6 % | 109× | 60.1 % |
Sonnet 5's cost is size-sensitive: a verbose model, its cost swings widely between short and long agentic tasks, hence its wide interval.
Fable 5.1vsHaiku 5.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5.1 | Haiku 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 70× | 64.8 % | 1.0× | 51.0 % |
| medium | 84× | 67.2 % | 1.4× | 55.5 % |
| high | 102× | 68.2 % | 2.2× | 58.6 % |
| xHigh | 186× | 69.8 % | 4.3× | 61.8 % |
| max | 293× | 70.6 % | 12× | 64.2 % |
Fable 5.1vsHaiku 4.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
Fable 5vsOpus 5.5
Opus 5.5 at Medium gives the same quality as Fable 5 at Max for 7.7× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5 | Opus 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 69× | 60.5 % | 25× | 66.0 % |
| medium | 95× | 64.0 % | 35× | 70.8 % |
| high | 115× | 65.9 % | 43× | 72.2 % |
| xHigh | 165× | 66.9 % | 74× | 73.9 % |
| max | 270× | 67.8 % | 194× | 74.7 % |
Fable 5vsOpus 5
Opus 5 at High gives the same quality as Fable 5 at Max for 3.2× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5 | Opus 5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 69× | 60.5 % | 39× | 62.5 % |
| medium | 95× | 64.0 % | 57× | 66.1 % |
| high | 115× | 65.9 % | 84× | 68.0 % |
| xHigh | 165× | 66.9 % | 111× | 69.2 % |
| max | 270× | 67.8 % | 148× | 69.6 % |
Fable 5vsSonnet 5.5
Sonnet 5.5 at xHigh gives the same quality as Fable 5 at Max for 8.2× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5 | Sonnet 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 69× | 60.5 % | 12× | 58.5 % |
| medium | 95× | 64.0 % | 13× | 61.7 % |
| high | 115× | 65.9 % | 18× | 66.9 % |
| xHigh | 165× | 66.9 % | 33× | 70.9 % |
| max | 270× | 67.8 % | 115× | 72.8 % |
Fable 5vsSonnet 5
Fable 5 at Low gives the same quality as Sonnet 5 at Max for 1.6× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5 | Sonnet 5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 69× | 60.5 % | 16× | 48.3 % |
| medium | 95× | 64.0 % | 23× | 52.8 % |
| high | 115× | 65.9 % | 33× | 55.9 % |
| xHigh | 165× | 66.9 % | 50× | 57.5 % |
| max | 270× | 67.8 % | 109× | 60.1 % |
Sonnet 5's cost is size-sensitive: a verbose model, its cost swings widely between short and long agentic tasks, hence its wide interval.
Fable 5vsHaiku 5.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5 | Haiku 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 69× | 60.5 % | 1.0× | 51.0 % |
| medium | 95× | 64.0 % | 1.4× | 55.5 % |
| high | 115× | 65.9 % | 2.2× | 58.6 % |
| xHigh | 165× | 66.9 % | 4.3× | 61.8 % |
| max | 270× | 67.8 % | 12× | 64.2 % |
Fable 5vsHaiku 4.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
Opus 5.5vsOpus 5
Opus 5.5 at Medium gives the same quality as Opus 5 at Max for 4.2× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Opus 5.5 | Opus 5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 25× | 66.0 % | 39× | 62.5 % |
| medium | 35× | 70.8 % | 57× | 66.1 % |
| high | 43× | 72.2 % | 84× | 68.0 % |
| xHigh | 74× | 73.9 % | 111× | 69.2 % |
| max | 194× | 74.7 % | 148× | 69.6 % |
Opus 5.5vsSonnet 5.5
Opus 5.5 at High gives the same quality as Sonnet 5.5 at Max for 2.6× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Opus 5.5 | Sonnet 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 25× | 66.0 % | 12× | 58.5 % |
| medium | 35× | 70.8 % | 13× | 61.7 % |
| high | 43× | 72.2 % | 18× | 66.9 % |
| xHigh | 74× | 73.9 % | 33× | 70.9 % |
| max | 194× | 74.7 % | 115× | 72.8 % |
Opus 5.5vsSonnet 5
Opus 5.5 at Low gives the same quality as Sonnet 5 at Max for 4.4× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Opus 5.5 | Sonnet 5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 25× | 66.0 % | 16× | 48.3 % |
| medium | 35× | 70.8 % | 23× | 52.8 % |
| high | 43× | 72.2 % | 33× | 55.9 % |
| xHigh | 74× | 73.9 % | 50× | 57.5 % |
| max | 194× | 74.7 % | 109× | 60.1 % |
Sonnet 5's cost is size-sensitive: a verbose model, its cost swings widely between short and long agentic tasks, hence its wide interval.
Opus 5.5vsHaiku 5.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Opus 5.5 | Haiku 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 25× | 66.0 % | 1.0× | 51.0 % |
| medium | 35× | 70.8 % | 1.4× | 55.5 % |
| high | 43× | 72.2 % | 2.2× | 58.6 % |
| xHigh | 74× | 73.9 % | 4.3× | 61.8 % |
| max | 194× | 74.7 % | 12× | 64.2 % |
Opus 5.5vsHaiku 4.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
Opus 5vsSonnet 5.5
Sonnet 5.5 at xHigh gives the same quality as Opus 5 at Max for 4.5× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Opus 5 | Sonnet 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 39× | 62.5 % | 12× | 58.5 % |
| medium | 57× | 66.1 % | 13× | 61.7 % |
| high | 84× | 68.0 % | 18× | 66.9 % |
| xHigh | 111× | 69.2 % | 33× | 70.9 % |
| max | 148× | 69.6 % | 115× | 72.8 % |
Opus 5vsSonnet 5
Opus 5 at Low gives the same quality as Sonnet 5 at Max for 2.8× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Opus 5 | Sonnet 5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 39× | 62.5 % | 16× | 48.3 % |
| medium | 57× | 66.1 % | 23× | 52.8 % |
| high | 84× | 68.0 % | 33× | 55.9 % |
| xHigh | 111× | 69.2 % | 50× | 57.5 % |
| max | 148× | 69.6 % | 109× | 60.1 % |
Sonnet 5's cost is size-sensitive: a verbose model, its cost swings widely between short and long agentic tasks, hence its wide interval.
Opus 5vsHaiku 5.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Opus 5 | Haiku 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 39× | 62.5 % | 1.0× | 51.0 % |
| medium | 57× | 66.1 % | 1.4× | 55.5 % |
| high | 84× | 68.0 % | 2.2× | 58.6 % |
| xHigh | 111× | 69.2 % | 4.3× | 61.8 % |
| max | 148× | 69.6 % | 12× | 64.2 % |
Opus 5vsHaiku 4.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
Sonnet 5.5vsSonnet 5
Sonnet 5.5 at Medium gives the same quality as Sonnet 5 at Max for 8.5× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Sonnet 5.5 | Sonnet 5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 12× | 58.5 % | 16× | 48.3 % |
| medium | 13× | 61.7 % | 23× | 52.8 % |
| high | 18× | 66.9 % | 33× | 55.9 % |
| xHigh | 33× | 70.9 % | 50× | 57.5 % |
| max | 115× | 72.8 % | 109× | 60.1 % |
Sonnet 5's cost is size-sensitive: a verbose model, its cost swings widely between short and long agentic tasks, hence its wide interval.
Sonnet 5.5vsHaiku 5.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Sonnet 5.5 | Haiku 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 12× | 58.5 % | 1.0× | 51.0 % |
| medium | 13× | 61.7 % | 1.4× | 55.5 % |
| high | 18× | 66.9 % | 2.2× | 58.6 % |
| xHigh | 33× | 70.9 % | 4.3× | 61.8 % |
| max | 115× | 72.8 % | 12× | 64.2 % |
Sonnet 5.5vsHaiku 4.5
Sonnet 5.5 at Low gives the same quality as Haiku 4.5 at its only setting for 1.1× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
Sonnet 5vsHaiku 5.5
Haiku 5.5 at xHigh gives the same quality as Sonnet 5 at Max for 25× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Sonnet 5 | Haiku 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 16× | 48.3 % | 1.0× | 51.0 % |
| medium | 23× | 52.8 % | 1.4× | 55.5 % |
| high | 33× | 55.9 % | 2.2× | 58.6 % |
| xHigh | 50× | 57.5 % | 4.3× | 61.8 % |
| max | 109× | 60.1 % | 12× | 64.2 % |
Sonnet 5's cost is size-sensitive: a verbose model, its cost swings widely between short and long agentic tasks, hence its wide interval.
Sonnet 5vsHaiku 4.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
Sonnet 5's cost is size-sensitive: a verbose model, its cost swings widely between short and long agentic tasks, hence its wide interval.
Haiku 5.5vsHaiku 4.5
Haiku 5.5 at Low gives the same quality as Haiku 4.5 at its only setting for 13× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
Every pair, written out
Fable 5.1vsOpus 5.5
Opus 5.5 at Medium gives the same quality as Fable 5.1 at Max for 8.3× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5.1 | Opus 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 70× | 64.8 % | 25× | 66.0 % |
| medium | 84× | 67.2 % | 35× | 70.8 % |
| high | 102× | 68.2 % | 43× | 72.2 % |
| xHigh | 186× | 69.8 % | 74× | 73.9 % |
| max | 293× | 70.6 % | 194× | 74.7 % |
Fable 5.1vsSonnet 5.5
Sonnet 5.5 at xHigh gives the same quality as Fable 5.1 at Max for 8.9× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5.1 | Sonnet 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 70× | 64.8 % | 12× | 58.5 % |
| medium | 84× | 67.2 % | 13× | 61.7 % |
| high | 102× | 68.2 % | 18× | 66.9 % |
| xHigh | 186× | 69.8 % | 33× | 70.9 % |
| max | 293× | 70.6 % | 115× | 72.8 % |
Fable 5.1vsHaiku 5.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5.1 | Haiku 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 70× | 64.8 % | 1.0× | 51.0 % |
| medium | 84× | 67.2 % | 1.4× | 55.5 % |
| high | 102× | 68.2 % | 2.2× | 58.6 % |
| xHigh | 186× | 69.8 % | 4.3× | 61.8 % |
| max | 293× | 70.6 % | 12× | 64.2 % |
Opus 5.5vsSonnet 5.5
Opus 5.5 at High gives the same quality as Sonnet 5.5 at Max for 2.6× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Opus 5.5 | Sonnet 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 25× | 66.0 % | 12× | 58.5 % |
| medium | 35× | 70.8 % | 13× | 61.7 % |
| high | 43× | 72.2 % | 18× | 66.9 % |
| xHigh | 74× | 73.9 % | 33× | 70.9 % |
| max | 194× | 74.7 % | 115× | 72.8 % |
Opus 5.5vsHaiku 5.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Opus 5.5 | Haiku 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 25× | 66.0 % | 1.0× | 51.0 % |
| medium | 35× | 70.8 % | 1.4× | 55.5 % |
| high | 43× | 72.2 % | 2.2× | 58.6 % |
| xHigh | 74× | 73.9 % | 4.3× | 61.8 % |
| max | 194× | 74.7 % | 12× | 64.2 % |
Sonnet 5.5vsHaiku 5.5
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Sonnet 5.5 | Haiku 5.5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 12× | 58.5 % | 1.0× | 51.0 % |
| medium | 13× | 61.7 % | 1.4× | 55.5 % |
| high | 18× | 66.9 % | 2.2× | 58.6 % |
| xHigh | 33× | 70.9 % | 4.3× | 61.8 % |
| max | 115× | 72.8 % | 12× | 64.2 % |
Fable 5.1vsFable 5
Fable 5.1 at Medium gives the same quality as Fable 5 at Max for 3.2× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Fable 5.1 | Fable 5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 70× | 64.8 % | 69× | 60.5 % |
| medium | 84× | 67.2 % | 95× | 64.0 % |
| high | 102× | 68.2 % | 115× | 65.9 % |
| xHigh | 186× | 69.8 % | 165× | 66.9 % |
| max | 293× | 70.6 % | 270× | 67.8 % |
Opus 5.5vsOpus 5
Opus 5.5 at Medium gives the same quality as Opus 5 at Max for 4.2× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Opus 5.5 | Opus 5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 25× | 66.0 % | 39× | 62.5 % |
| medium | 35× | 70.8 % | 57× | 66.1 % |
| high | 43× | 72.2 % | 84× | 68.0 % |
| xHigh | 74× | 73.9 % | 111× | 69.2 % |
| max | 194× | 74.7 % | 148× | 69.6 % |
Sonnet 5.5vsSonnet 5
Sonnet 5.5 at Medium gives the same quality as Sonnet 5 at Max for 8.5× less per task.
By effort level cost per task × the cheapest couple · expected score on the benchmark panel
| Effort | Sonnet 5.5 | Sonnet 5 | ||
|---|---|---|---|---|
| cost | score | cost | score | |
| low | 12× | 58.5 % | 16× | 48.3 % |
| medium | 13× | 61.7 % | 23× | 52.8 % |
| high | 18× | 66.9 % | 33× | 55.9 % |
| xHigh | 33× | 70.9 % | 50× | 57.5 % |
| max | 115× | 72.8 % | 109× | 60.1 % |
Sonnet 5's cost is size-sensitive: a verbose model, its cost swings widely between short and long agentic tasks, hence its wide interval.
How to read these comparisons
The values are the ones the main page shows, fitted from public measurements taken on the same tasks. One model is cheaper or higher than the other at an effort level when the fitted difference puts it ahead with at least 84 % probability, the level of the intervals shown everywhere on the site; otherwise the two are level within the uncertainty. A match is the cheapest setting of the other model that the first does not out-score at that level. Costs are what a whole task cost, as each source measured it (the run's actual spend, cache included), not the price per token. So the cost already counts how many tokens each model uses: a model with a lower price per token can still cost more per task if it writes many more tokens.