101 languages, one embedding space
No translation step, no per-language index, and no additional resident memory. A query in any supported language retrieves relevant passages in any other.
The 30 highest-volume languages
These have their own reference pages with sample queries. The lang filter accepts the full ISO set regardless.
English
Englishenclimate policy
Spanish
Españolespolítica climática
Chinese
中文zh气候政策
German
DeutschdeKlimapolitik
French
Françaisfrpolitique climatique
Japanese
日本語ja気候政策
Portuguese
Portuguêsptpolítica climática
Russian
Русскийruклиматическая политика
Italian
Italianoitpolitica climatica
Dutch
Nederlandsnlklimaatbeleid
Korean
한국어ko기후 정책
Arabic
العربيةarسياسة المناخ
Hindi
हिन्दीhiजलवायु नीति
Polish
Polskiplpolityka klimatyczna
Turkish
Türkçetriklim politikası
Swedish
Svenskasvklimatpolitik
Indonesian
Bahasa Indonesiaidkebijakan iklim
Vietnamese
Tiếng Việtvichính sách khí hậu
Ukrainian
Українськаukкліматична політика
Czech
Češtinacsklimatická politika
Romanian
Românăropolitica climatică
Greek
Ελληνικάelκλιματική πολιτική
Hebrew
עבריתheמדיניות אקלים
Danish
Danskdaklimapolitik
Finnish
Suomifiilmastopolitiikka
Norwegian
Norsknoklimapolitikk
Hungarian
Magyarhuklímapolitika
Thai
ไทยthนโยบายสภาพภูมิอากาศ
Bengali
বাংলাbnজলবায়ু নীতি
Persian
فارسیfaسیاست اقلیمی
Frequently asked questions
How many languages does unlob support?
101, all in a single embedding space. The lang filter accepts the full ISO set; the 30 highest-volume languages have their own reference pages.
Do I need to translate my query?
No. Semantic and hybrid modes cross languages natively, because texts with the same meaning land near each other in the shared space regardless of language. Translating the query would search for the translation rather than the meaning.
Does keyword search work across languages?
No — keyword mode matches tokens, and an English token does not appear in a German document. Use semantic or hybrid mode for cross-language retrieval. Hybrid, the default, gives you lexical precision within the query language and semantic reach across all of them.
How are Chinese, Japanese and Thai handled?
As character bigrams, because those scripts do not delimit words with spaces. 量子力学 indexes as 量子, 子力 and 力学, with the same rule applied to queries so the two always agree. It needs no language-specific segmentation model.
Is coverage equal across languages?
No, and it would be misleading to imply otherwise — English is by far the largest slice of the corpus. Set facets=true to see the language distribution of a result set before you answer, rather than discovering thin coverage from a user complaint.
What one embedding space removes
The conventional multilingual stack puts a translation step in front of a per-language index. It carries two costs that compound: infrastructure multiplied by the number of languages, and meaning lost at the translation boundary, because the query that actually runs is a translation of the question rather than the question. The second failure is the dangerous one, since something always comes back and nothing signals that it is the wrong something.
Placing every language in one space removes the step rather than improving it. A question asked in one language retrieves a passage written in another because the passage is about the thing that was asked, and no translation was involved in deciding that.
One space is not equal coverage
Shared retrieval does not mean shared volume. English is by far the largest slice of the open web and therefore of any index over it, so a cross-language query will lean on it unless told otherwise. That is frequently the right outcome and occasionally a serious problem — a question about local regulation answered from English secondary coverage is a worse answer than a thinner set of primary sources in the language they were written in.
So look before assuming. A probe that returns counts and language distribution with no results is cheap, and it lets a product say coverage is thin here instead of silently returning weaker sources. Pin the language when the primary source has to be in it; leave it unpinned when meaning is what matters. Themultilingual guide goes through both cases.
Search every language from one query
10,000 free credits a month, no card. All languages are available on every plan.