DebateCardAI vs haku.cards: Which Debate Tool Should You Use?
Prices and features checked July 28, 2026. We build DebateCardAI.
haku is a search engine over cards other debaters already cut, and it is a good one. Their index is 3.5 million cards from openCaselist, 2013 to 2026, across hspolicy, hsld, nfald, hspf, and ndtceda.
We have that too. Ours is called Logos, it holds about a million cards from the same wiki, and it is free on every account. The difference is that for haku the search is the product, and for us it is one tool of about a dozen.
Our position, plainly: the index size difference is mostly years, and we think years are the wrong thing to buy. Topics repeat, so the same card gets recut and redisclosed season after season, and a 2013-2026 index means more near-identical versions of the card you already found, not more coverage. Logos stays focused on the recent seasons where the live evidence is, re-ingests weekly, and spends the effort on the search itself: free and uncapped, finds arguments rather than keywords, and works from inside Claude. And once you have the card, we think nothing else in this category comes close to what we do next with it.
So the real question is whether you need a better front end on openCaselist, or the cards that are not on openCaselist yet.
Feature by feature
| haku.cards | DebateCardAI | |
|---|---|---|
| Wiki card search | 3.5M indexed | 1M+, free |
| Filter by tournament and round | yes | no |
| Filter by year, event, school, side | yes | yes |
| Search by argument, not keyword | no | Deep Search |
| Pull a whole A2 block at once | no | block search |
| Search from inside Claude | no | MCP connector |
| Card cutter | yes | three modes |
| Cut from the live web | no | Evidence Finder |
| Condense cards for reading | no | Rhetorify |
| Write a full speech | no | Speechify |
| Build A2 blocks | no | Block Builder |
| Judge prep | Judge Prefs | no |
| Tournament math | no | Bracket Math |
| Look up a debater | no | Debater Search |
| Chrome extension | no | yes |
| Pricing | not published | free tier, $8.99/mo |
What haku does better
Their filters. You can narrow by year, event, school, team, side, tournament, and round, and the last two are the ones we do not have. If the job is "find the neg card this specific team read in round 4 at this specific tournament," haku gets you there and we do not.
Their index is also larger, and they are honest about what it is. Both of our numbers describe openCaselist, the public wiki where teams disclose what they read, maintained by Aaron Hardy alongside Verbatim. Neither of us has a secret stash, and most of the big card counts in this category are the same wiki counted differently.
They also ship a Judge Prefs tool. We do not do judge prep at all. Our Bracket Math and Debater Search are tournament-day tools about the other team, not about the person voting, so they are not a substitute.
One thing we could not find: haku does not publish pricing on any public page, so we cannot tell you what it costs. I would not assume free just because there is no paywall in front of the search box.
Where we go further
Searching by argument. Keyword search only finds the phrasing you guessed, and the same argument gets tagged a dozen ways. Deep Search takes the argument as you would say it ("I need a card saying deterrence prevents war") and finds cards that make it regardless of wording, with a line on each result explaining the match. You can also pull an entire answers-to file with one block-header search instead of collecting cards one at a time.
Cards nobody has cut. The wiki is a ceiling. Evidence Finder takes a claim with no source behind it, searches the web and academic databases, ranks results by how cuttable they are, and cuts the card. Those are also the cards your opponent has never seen.
Cutting. One card from a URL, a batch you approve as it lands, or a queue you start and review at the end, with style and highlight density as settings so the card comes out shaped for your panel. Cutting is the part we have built hardest, and if you want a card cut to your argument we think we are the best option in this category. Test it on one article and you will know.
After you find it. Rhetorify strips cards to what you say out loud, in your browser, no AI, nothing metered. Speechify writes the whole constructive or rebuttal from the stack. Block Builder walks from a strategy chat to a finished block. Search to card to speech in one place is the pipeline, and it is the strongest reason we would point a searcher our way.
In Claude. Our connector lets Claude search the card library mid-conversation and hand back real disclosed cards with source URLs.
Which one
Pick haku if your prep runs on the wiki and you want the deepest tournament and round filters available.
Pick DebateCardAI if you need cards nobody has cut, if you want search that understands the argument rather than the wording, or if you want cutting, speech, and block prep connected to the search.
Both works fine. Search theirs for what exists, cut ours for what does not.
Switching
Nothing to migrate; cards are text. Paste them into our library or Rhetorify and they work.
The fastest test is to run the same argument through both searches and see which one hands you the card you were actually looking for. Takes five minutes and beats reading either of our marketing pages.
Try DebateCardAI free or read the 2026 tool guide.