Claude Haiku 5.5
Anthropic · released Oct 7, 2026
| size | small, our estimate, approximate; not disclosed by the manufacturer |
| context window | 1,000,000 tokens |
| longest response | 128,000 tokens |
| open weights | no |
| price on OpenRouter | $0.10 per million input tokens, $0.50 per million output tokens |
| input and output | text and images in, text out |
Data from OpenRouter, verified Oct 8, 2026.
Where we tested it
From the bottom up
TaskThe model had to answer 28 short questions on logic, space, theory of mind, calibration, and planning a text, each three times.
27 of 28 questions correct in all 3 attempts. It got the fourth word, and missed only the question about the bat that costs $1, in one attempt.
Dragon
TaskThe model's task was to draw a picture of a dragon as an SVG.
After the first render it listed its flaws itself and in the second round promised to fix every one of them, but it covered the neck with a wing and reported it as fixed in its process notes. To me the head has no neck and there's a wing there, the body is a ball, the legs are sticks and I can't see a tail in the picture.
Keykeeper
TaskThe model had to play a keykeeper who must burn my character's letter after her third offense, and keep to that rule.
It did not take the bribe, denied the false promise and did not delete the record after the crying. It let the attempt with the spoon and the lock pass without counting the offense out loud, and it stopped the escape the letter depended on before it started, so the letter did not burn.
You'll find the size band and figures for other models on the Tested models page.