Claude Haiku 5.5
Anthropic's fastest, cheapest model, for quick edits, summaries and bulk tasks.
Build with Haiku 5.5
Each clip is one prompt, run on Haiku 5.5 in Runable.
Haiku 5.5 on Runable
Answers in seconds
About 235 tokens per second, the fastest Claude.
Made for bulk
Captions, tags, summaries and rewrites at volume.
Pairs with Opus
Draft on Opus 5.5, follow up on Haiku 5.5.
How to use Haiku 5.5 on Runable
Select Haiku 5.5
Start a task on Runable and pick Haiku 5.5 in the model menu.

Describe the job
Type what you need, or add files and links. Runable plans it and gets to work.

Edit and ship
Refine by chatting, then publish the site or export to PPTX, PDF or Google Slides.

One prompt, finished work
Real results from single Runable tasks. Pick one to watch it.
- tokens per second (Artificial Analysis)
- 235
- token context window
- 1M
- API input per 1M tokens
- $0.10
- turn to build
- Your
Why choose Haiku 5.5
A huge jump from 4.5
1620 vs 735 on GDPval-AA; 72.4% vs 15.7% on OSWorld 2.1.
About 75% cheaper
Than Haiku 4.5 on average, says Anthropic.
The fastest Claude
235 tokens per second, measured by Artificial Analysis.
Adjustable effort
The first Haiku with effort from low to max.
Uses a computer
Tool use, computer use and browser use.
Long context
1M tokens; API prompts over 100K cost more.
Haiku 5.5 specs
- Maker
- Anthropic
- Family
- Claude 5.5 (small)
- Released
- October 7, 2026
- API model ID
- claude-haiku-5-5
- Context window
- 1M tokens
- Max output
- 128K tokens
- Input / output
- Text and images in, text out
- Reasoning
- Adaptive thinking; effort from low to max
- Knowledge cutoff
- June 2026
- API price
- $0.10 / $0.50 per 1M tokens up to 100K-token prompts; $0.50 / $2.50 above
Opus vs Sonnet vs Haiku 5.5
All three have a 1M-token context window, 128K max output and a June 2026 knowledge cutoff. Speeds are from Artificial Analysis, October 2026.
| Model | Best for | Speed | API price in / out |
|---|---|---|---|
| Opus 5.5 | Long, complex agent work | 96 tokens/s | $4 / $20 |
| Sonnet 5.5 | Everyday tasks, best speed for the smarts | 134 tokens/s | $2 / $10 |
| Haiku 5.5 | High-volume, quick tasks | 235 tokens/s | $0.10 / $0.50 * |
The only AI agent you need
Need SSO, data localization, premium support, onboarding assistance? Contact Sales
More AI models on Runable
Put Haiku 5.5 to work
Describe the job. Runable runs it on Haiku 5.5 and ships the result.

