Spider for AI agents: benchmark results
Spider (content API) was tested on scrape, not on search, map and crawl. It got the same tasks, the same model (Claude Haiku 5.5), the same prompt and the same limits as every other provider, and a script checked every answer. Spider was tested on scrape only, because its free wallet was small.
Scrape 60 tasks · Read one page and pull out exact fields: prices, versions, install commands.
- Tasks done
- 20%
- 12 of 60
- 95% interval
- 4%–100%
- Cost per finished task
- 2.48¢
- Median time
- 4.6 s
8th of 8 on tasks done. Spider is behind Context.dev (92%), but the gap of 72 points is too small to rule out chance.
Raw output, no model: the API's own response held every answer on 10 of 55 applicable tasks (18%), at 0.04¢ of API spend per finished task.
| Provider | Tasks done | 95% interval | Cost per finished task | Median time |
|---|---|---|---|---|
| Context.dev | 55 of 60 (92%) | 90%–100% | 0.72¢ | 3.7 s |
| Linkup | 54 of 60 (90%) | 90%–90% | 0.60¢ | 3.4 s |
| Firecrawl | 54 of 60 (90%) | 90%–90% | 1.22¢ | 3.3 s |
| TinyFish | 54 of 60 (90%) | 80%–92% | 0.43¢ | 3.5 s |
| Scrapfly | 53 of 60 (88%) | 80%–90% | 0.61¢ | 5.0 s |
| Jina | 52 of 60 (87%) | 60%–92% | 0.57¢ | 3.8 s |
| Browserbase | 49 of 60 (82%) | 78%–100% | 1.12¢ | 3.3 s |
| Spider | 12 of 60 (20%) | 4%–100% | 2.48¢ | 4.6 s |
Read these numbers with
- Spider refused our test site because its malicious-URL filter flags the workers.dev domain. Three inspected responses carry that explicit 403.
- It scraped all 10 PyPI pages correctly. Two of its other passes are tasks where the right answer was empty, which it returned because the page was blocked.
- This is a domain compatibility finding, not a measure of how well Spider scrapes sites it accepts.
Spider compared
- Firecrawl vs Spider scrape, compared with Firecrawl