Skip to content
Wycord
Research

AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African Languages vs IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models

A verified, side-by-side comparison. Both records are status-checked by Wycord, so you are comparing what each actually offers today, not a stale listing.

AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African LanguagesIrokoBench: A New Benchmark for African Languages in the Age of Large Language Models
Category
Research
Research
Type
NLP benchmark
NLP benchmark
Country
🌍 Pan-African
🌍 Pan-African
Docs status
Docs live
Docs live
Licensing
Pricing
Free / open
Free / open
Verified
Unverified
Unverified
Last verified
10 Oct 2026
10 Oct 2026
Tags
african-languages, hate-speech, nlp-benchmark, content-moderation
african-languages, masakhane, nlp-benchmark, llm-evaluation
Links
Website
Website
Summary
AfriHate is a multilingual benchmark of hate speech and abusive language datasets covering 15 African languages, annotated by native speakers. The paper contributes classification baselines and hate speech and offensive language lexicons, and analyses why keyword-based moderation fails for low-resource African languages. It was released on arXiv in January 2025.
IrokoBench is a human-translated evaluation benchmark covering 17 typologically diverse low-resource African languages across three tasks: natural language inference (AfriXNLI), mathematical reasoning (AfriMGSM) and knowledge-based multiple-choice QA (AfriMMLU). The paper evaluates open and proprietary LLMs and documents a large gap between high-resource languages and African languages, with the best open model reaching about 63 percent of GPT-4o performance. It was published at NAACL 2025.

AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African Languages vs IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models at a glance

AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African Languages and IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models are both listed under NLP benchmark in Wycord's Research index.

Both are based in Pan-African.

Both currently show the status "Docs live".

On access, both are marked "No approval needed".

Over the last 30 days Wycord's probes measured 100% uptime for AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African Languages (36 checks) and 100% for IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models (36 checks).

Average response time on live probes was 6 ms for AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African Languages and 25 ms for IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models.

Uptime and latency come from Wycord's own probes of each resource's website, documentation and repository, run every six hours.

How they differ

  • Average latency: AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African Languages, 6 ms; IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models, 25 ms.

Topics

Both are tagged: african-languages, nlp-benchmark.

Only AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African Languages: hate-speech, content-moderation.

Only IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models: masakhane, llm-evaluation.

Common questions

Which has better uptime, AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African Languages or IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models?
They are level at 100% over the last 30 days.
Do AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African Languages or IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models need approval before use?
Both are marked "No approval needed".
Is the documentation for AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African Languages and IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models working?
Both show "Docs live" at the last check.
Where are AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African Languages and IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models based?
Both are based in Pan-African.

More comparisons

Full details: AfriHate: A Multilingual Collection of Hate Speech and Abusive Language Datasets for African LanguagesFull details: IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models