INJONGO
Models identify a speaker’s intent and extract information from utterances in 16 African languages.
AfroBench uses the intent detection component.
- Made by
- Hao Yu and colleagues
- Released
- February 2025
- Kind
- Reading in many languages
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- All public
- Measures
- Capability
- Language
- Many languages
- Interaction
- Single turn
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Many languages
- Strongest model tested
- GPT-4o Aug 2024 #167 on the capability index
- Last model added
- December 2024
- Scores refreshed
- November 20th, 2026
AfroXLMR-76L tops the leaderboard with 94.4% GPT-4o leads with 65.1% among models released by December 2024
Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. The headline is mean intent detection accuracy across the 16 African languages. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is GPT-4o, at 65.1%. This view shows only models we track, each at its best.
The newest model with a score here came out in December 2024, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1 AfroXLMR-76L Multilingual training 94.4%
- 2 AfroXLMR 76L In-language training 93.7%
- 3 NLLB LLM2Vec In-language training 93.4%
- 4 AfroXLMR In-language training 92.2%
- 5 AfriTeVa V2 (T5) In-language training 91.7%
- 6 AfriBERTa V2 In-language training 88.6%
- 7 mT5-Large In-language training 87.7%
- 8 XLM-RoBERTa In-language training 84.5%
- 9 GPT-4o Aug 2024, Zero-shot 65.1%
- 10 Gemini 1.5 Pro (snapshot unspecified) Zero-shot 59.1%
- 11 Aya-101 13B Zero-shot 42.4%
- 12
Llama 3.3 70B Instruct Zero-shot 38.5%
- 13
Gemma 2 27B Zero-shot 33%
- 14
Gemma 2 9B Zero-shot 29.2%
- 15
Llama 3.1-8B Zero-shot 3.90%
Source: The INJONGO paper
- 1 GPT-4o Aug 2024, Zero-shot 65.1%
- 2
Llama 3.3 70B Instruct Zero-shot 38.5%
- 3
Gemma 2 27B Zero-shot 33%
- 4
Gemma 2 9B Zero-shot 29.2%
- 5
Llama 3.1-8B Zero-shot 3.90%
Source: The INJONGO paper
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | AfroXLMR-76L | 94.4% | Multilingual training | Official leaderboard | |
| 2 | AfroXLMR 76L | 93.7% | In-language training | Official leaderboard | |
| 3 | NLLB LLM2Vec | 93.4% | In-language training | Official leaderboard | |
| 4 | AfroXLMR | 92.2% | In-language training | Official leaderboard | |
| 5 | AfriTeVa V2 (T5) | 91.7% | In-language training | Official leaderboard | |
| 6 | AfriBERTa V2 | 88.6% | In-language training | Official leaderboard | |
| 7 | mT5-Large | 87.7% | In-language training | Official leaderboard | |
| 8 | XLM-RoBERTa | 84.5% | In-language training | Official leaderboard | |
| 9 | GPT-4o Aug 2024 | 65.1% | Zero-shot | Aug 2024 | Official leaderboard |
| 10 | Gemini 1.5 Pro (snapshot unspecified) | 59.1% | Zero-shot | Official leaderboard | |
| 11 | Aya-101 13B | 42.4% | Zero-shot | Official leaderboard | |
| 12 | | 38.5% | Zero-shot | Dec 2024 | Official leaderboard |
| 13 | | 33% | Zero-shot | Jun 2024 | Official leaderboard |
| 14 | | 29.2% | Zero-shot | Jun 2024 | Official leaderboard |
| 15 | | 3.90% | Zero-shot | Jul 2024 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-4o Aug 2024 | 65.1% | Zero-shot | Aug 2024 | Official leaderboard |
| 2 | | 38.5% | Zero-shot | Dec 2024 | Official leaderboard |
| 3 | | 33% | Zero-shot | Jun 2024 | Official leaderboard |
| 4 | | 29.2% | Zero-shot | Jun 2024 | Official leaderboard |
| 5 | | 3.90% | Zero-shot | Jul 2024 | Official leaderboard |
Rankings by part
We read 35 parts from the benchmark’s paper in 2 groups: intent detection and slot filling. Each ranks models on one part of the test. GPT-4o leads 32 of them, and Llama 3.3 70B Instruct leads the other 3. Arrows show how far a model moved from its overall rank among the models we track.
eng
5 models we track have a score here.
- 1
Llama 3.3 70B Instruct ↑1 81.1%
- 2 GPT-4o Aug 2024 ↓1 80.9%
- 3
Gemma 2 27B 79.5%
- 4
Gemma 2 9B 77.6%
- 5
Llama 3.1-8B 27.6%
amh
5 models we track have a score here.
- 1 GPT-4o Aug 2024 76%
- 2
Llama 3.3 70B Instruct 56.2%
- 3
Gemma 2 9B ↑1 49.2%
- 4
Gemma 2 27B ↓1 47.2%
- 5
Llama 3.1-8B 1.90%
ewe
5 models we track have a score here.
- 1 GPT-4o Aug 2024 15.1%
- 2
Llama 3.3 70B Instruct 9.50%
- 3
Gemma 2 27B 6.30%
- 4
Gemma 2 9B 6.10%
- 5
Llama 3.1-8B 2.10%
hau
5 models we track have a score here.
- 1 GPT-4o Aug 2024 80.7%
- 2
Llama 3.3 70B Instruct 52.3%
- 3
Gemma 2 27B 46.5%
- 4
Gemma 2 9B 40.8%
- 5
Llama 3.1-8B 4.80%
ibo
5 models we track have a score here.
- 1 GPT-4o Aug 2024 71.8%
- 2
Llama 3.3 70B Instruct 52.4%
- 3
Gemma 2 27B 36.9%
- 4
Gemma 2 9B 31.5%
- 5
Llama 3.1-8B 5.50%
kin
5 models we track have a score here.
- 1 GPT-4o Aug 2024 64.7%
- 2
Llama 3.3 70B Instruct 35%
- 3
Gemma 2 27B 26.7%
- 4
Gemma 2 9B 23.8%
- 5
Llama 3.1-8B 3.30%
lin
5 models we track have a score here.
- 1 GPT-4o Aug 2024 56.4%
- 2
Llama 3.3 70B Instruct 37.5%
- 3
Gemma 2 27B 27.5%
- 4
Gemma 2 9B 22.2%
- 5
Llama 3.1-8B 5.30%
lug
5 models we track have a score here.
- 1 GPT-4o Aug 2024 68.2%
- 2
Llama 3.3 70B Instruct 37.7%
- 3
Gemma 2 27B 26.1%
- 4
Gemma 2 9B 23.2%
- 5
Llama 3.1-8B 2.40%
orm
5 models we track have a score here.
- 1 GPT-4o Aug 2024 59.3%
- 2
Llama 3.3 70B Instruct 12.4%
- 3
Gemma 2 9B ↑1 7.70%
- 4
Gemma 2 27B ↓1 5.80%
- 5
Llama 3.1-8B 1.60%
sna
5 models we track have a score here.
- 1 GPT-4o Aug 2024 75.5%
- 2
Gemma 2 27B ↑1 36.7%
- 3
Llama 3.3 70B Instruct ↓1 32.3%
- 4
Gemma 2 9B 29.7%
- 5
Llama 3.1-8B 2.80%
sot
5 models we track have a score here.
- 1 GPT-4o Aug 2024 59.7%
- 2
Llama 3.3 70B Instruct 30.5%
- 3
Gemma 2 27B 25.6%
- 4
Gemma 2 9B 19.9%
- 5
Llama 3.1-8B 2.90%
swa
5 models we track have a score here.
- 1 GPT-4o Aug 2024 84.5%
- 2
Llama 3.3 70B Instruct 80.6%
- 3
Gemma 2 27B 75.5%
- 4
Gemma 2 9B 70%
- 5
Llama 3.1-8B 14.1%
twi
5 models we track have a score here.
- 1 GPT-4o Aug 2024 58.6%
- 2
Llama 3.3 70B Instruct 29.3%
- 3
Gemma 2 27B 21.2%
- 4
Gemma 2 9B 21%
- 5
Llama 3.1-8B 2.60%
wol
5 models we track have a score here.
- 1 GPT-4o Aug 2024 43.7%
- 2
Llama 3.3 70B Instruct 20.9%
- 3
Gemma 2 27B 16.4%
- 4
Gemma 2 9B 13.8%
- 5
Llama 3.1-8B 4.00%
xho
5 models we track have a score here.
- 1 GPT-4o Aug 2024 79.6%
- 2
Gemma 2 27B ↑1 50.2%
- 3
Llama 3.3 70B Instruct ↓1 43.5%
- 4
Gemma 2 9B 40.1%
- 5
Llama 3.1-8B 3.20%
yor
5 models we track have a score here.
- 1 GPT-4o Aug 2024 77%
- 2
Llama 3.3 70B Instruct 41.4%
- 3
Gemma 2 27B 34.8%
- 4
Gemma 2 9B 32.2%
- 5
Llama 3.1-8B 3.50%
zul
5 models we track have a score here.
- 1 GPT-4o Aug 2024 71.2%
- 2
Gemma 2 27B ↑1 44.3%
- 3
Llama 3.3 70B Instruct ↓1 43.9%
- 4
Gemma 2 9B 36.3%
- 5
Llama 3.1-8B 2.80%
eng
5 models we track have a score here.
- 1 GPT-4o Aug 2024 55.4%
- 2
Llama 3.3 70B Instruct 52.6%
- 3
Gemma 2 27B 49.8%
- 4
Gemma 2 9B 34.1%
- 5
Llama 3.1-8B 25%
amh
5 models we track have a score here.
- 1
Llama 3.3 70B Instruct ↑1 26.3%
- 2 GPT-4o Aug 2024 ↓1 22.8%
- 3
Gemma 2 27B 15.7%
- 4
Gemma 2 9B 4.50%
- 5
Llama 3.1-8B 3.70%
ewe
5 models we track have a score here.
- 1
Llama 3.3 70B Instruct ↑1 22%
- 2 GPT-4o Aug 2024 ↓1 19.4%
- 3
Gemma 2 27B 9.50%
- 4
Llama 3.1-8B ↑1 5.60%
- 5
Gemma 2 9B ↓1 0.30%
hau
5 models we track have a score here.
- 1 GPT-4o Aug 2024 37.8%
- 2
Llama 3.3 70B Instruct 29.5%
- 3
Gemma 2 27B 24.1%
- 4
Llama 3.1-8B ↑1 11.1%
- 5
Gemma 2 9B ↓1 7.40%
ibo
5 models we track have a score here.
- 1 GPT-4o Aug 2024 38.9%
- 2
Llama 3.3 70B Instruct 35%
- 3
Gemma 2 27B 25.2%
- 4
Llama 3.1-8B ↑1 12.6%
- 5
Gemma 2 9B ↓1 10.6%
kin
5 models we track have a score here.
- 1 GPT-4o Aug 2024 36.4%
- 2
Llama 3.3 70B Instruct 31.4%
- 3
Gemma 2 27B 21.7%
- 4
Llama 3.1-8B ↑1 8.50%
- 5
Gemma 2 9B ↓1 5.00%
lin
5 models we track have a score here.
- 1 GPT-4o Aug 2024 33.5%
- 2
Llama 3.3 70B Instruct 25%
- 3
Gemma 2 27B 15.2%
- 4
Llama 3.1-8B ↑1 9.10%
- 5
Gemma 2 9B ↓1 6.00%
lug
5 models we track have a score here.
- 1 GPT-4o Aug 2024 35.3%
- 2
Llama 3.3 70B Instruct 30.4%
- 3
Gemma 2 27B 28.4%
- 4
Llama 3.1-8B ↑1 10.1%
- 5
Gemma 2 9B ↓1 5.60%
orm
5 models we track have a score here.
- 1 GPT-4o Aug 2024 13%
- 2
Llama 3.3 70B Instruct 9.30%
- 3
Llama 3.1-8B ↑2 2.80%
- 4
Gemma 2 27B ↓1 2.60%
- 5
Gemma 2 9B ↓1 0.10%
sna
5 models we track have a score here.
- 1 GPT-4o Aug 2024 40.2%
- 2
Gemma 2 27B ↑1 29.8%
- 3
Llama 3.3 70B Instruct ↓1 29.5%
- 4
Llama 3.1-8B ↑1 9.90%
- 5
Gemma 2 9B ↓1 7.30%
sot
5 models we track have a score here.
- 1 GPT-4o Aug 2024 40.9%
- 2
Llama 3.3 70B Instruct 36.4%
- 3
Gemma 2 27B 28%
- 4
Llama 3.1-8B ↑1 11.5%
- 5
Gemma 2 9B ↓1 10.8%
swa
5 models we track have a score here.
- 1 GPT-4o Aug 2024 46.5%
- 2
Llama 3.3 70B Instruct 40.7%
- 3
Gemma 2 27B 40.2%
- 4
Gemma 2 9B 21.2%
- 5
Llama 3.1-8B 17.3%
twi
5 models we track have a score here.
- 1 GPT-4o Aug 2024 40.1%
- 2
Llama 3.3 70B Instruct 35.6%
- 3
Gemma 2 27B 24.3%
- 4
Llama 3.1-8B ↑1 11.2%
- 5
Gemma 2 9B ↓1 2.40%
wol
5 models we track have a score here.
- 1 GPT-4o Aug 2024 37.9%
- 2
Llama 3.3 70B Instruct 36.4%
- 3
Gemma 2 27B 23.3%
- 4
Llama 3.1-8B ↑1 9.20%
- 5
Gemma 2 9B ↓1 2.60%
xho
5 models we track have a score here.
- 1 GPT-4o Aug 2024 10%
- 2
Llama 3.3 70B Instruct 6.90%
- 3
Gemma 2 27B 4.50%
- 4
Llama 3.1-8B ↑1 2.60%
- 5
Gemma 2 9B ↓1 2.20%
yor
5 models we track have a score here.
- 1 GPT-4o Aug 2024 42.4%
- 2
Llama 3.3 70B Instruct 34.2%
- 3
Gemma 2 27B 28.1%
- 4
Llama 3.1-8B ↑1 11%
- 5
Gemma 2 9B ↓1 5.20%
zul
5 models we track have a score here.
- 1 GPT-4o Aug 2024 37.6%
- 2
Llama 3.3 70B Instruct 31.9%
- 3
Gemma 2 27B 31%
- 4
Llama 3.1-8B ↑1 9.00%
- 5
Gemma 2 9B ↓1 8.20%
Slot filling F1
5 models we track have a score here.
- 1 GPT-4o Aug 2024 33.3%
- 2
Llama 3.3 70B Instruct 28.8%
- 3
Gemma 2 27B 22%
- 4
Llama 3.1-8B ↑1 9.10%
- 5
Gemma 2 9B ↓1 6.20%
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: The INJONGO paper
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official INJONGO leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Multilingual
- MMMLU In use MMLU professionally translated into 14 languages. Multiple-choice accuracy. Claude Mythos Preview 92.7%
- WMT24++ In use English-to-55-language translation of WMT24 test data. Scored with automatic metrics such as MetricX or xCOMET.
K-EXAONE 90.5%
- MMLU-ProX In use MMLU-Pro translated into up to 29 languages with expert review. Multiple-choice accuracy per language.
Qwen3.7-Max 87%
- CLIcK In use Models answer 1,995 multiple-choice questions on Korean language and culture. Questions come from official Korean exams and textbooks in eleven categories.
Solar Open 2 90.7%
- Global MMLU Lite 3.0 In use Models answer 400 human-translated MMLU questions per language in 23 languages. Version 3.0 adds Czech, Hungarian, Oriya, Slovak and Tajik and updates Italian. Accuracy. Claude Opus 5 95.7%
- MMMB In use Models answer multiple-choice questions about images in six languages across 15 categories. The set has about 12,000 questions. LFM2.5-VL-3B 83%
Data audit
9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 23 months ago, in December 2024, for Llama 3.3 70B Instruct. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
- Pass Independent runs Scores from the benchmark’s paper (15 entries).
- Pass Models scored 5 models we track, plus 10 other entries on its leaderboard.
- Warning Newest score The newest came 23 months ago, in December 2024, for Llama 3.3 70B Instruct.
- Note Frontier models None of the top 10 has a score. The strongest tested is GPT-4o (#167).
- Pass Headroom The best is 65.1%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on aclanthology.org: 5 rows for models we track and 10 other entries at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.