Bambara - UNICEF WCARO NLP

🌍 General Information

ISO 639-3bam

ISO 639-1bm

French Namebambara

Glottocodebamb1269

Language FamilyNiger–Congo

Writing SystemLatin script Arabic script ( Ajami ) NKo script Masaba [ 2 ]

L1 Speakers4.2 million

L2 Speakers10 million

Population10000000 (10⁶+)

Endangermentsafe

CountriesBurkina Faso, Cote d'Ivoire, Gambia, Ghana, Guinea, Mali, Mauritania, Niger, Senegal

Also Known AsBamako, Bamana, Bamanakan

WikipediaBambara on Wikipedia

🗺️ Language Use

Data from CLEAR Global's Language Use Data Platform.

👥 Actors (5)

Farmerline / Darli AI (Startup)

Ghanaian agritech company building technology solutions for smallholder farmers. Core platform Mergdata is a cloud-based farm management system used b...

MALIBA-AI (Community Initiative)

MALIBA-AI ("L'IA au service du Mali" - AI in service of Mali) is a community-driven organization focused on empowering Mali through open-source AI inn...

ML Collective (Ngo)

Non-profit research lab providing research opportunities to independent and underrepresented researchers. Runs multiple African language NLP projects ...

RobotsMali (Ngo)

RobotsMali is a centre dedicated to training and innovation in robotics and artificial intelligence, working to democratize access to emerging technol...

UBC Deep Learning & NLP Lab (University)

Research lab at the University of British Columbia led by Muhammad Abdul-Mageed, focused on deep representation learning and natural language socio-pr...

🤖 NLP & Tech Resources

✓ AfroLID

✓ Google Translate: 2022

✓ NLLB-200

✓ Mozilla Common Voice: ☑️

✓ MADLAD-400 Docs: 702 MAD-Docs

✓ MADLAD-400 Sentences: 24500 MAD-Sents

✓ Fineweb2: 💬

✓ CMU Wilderness

✓ MAFAND

✓ MMS-ASR

✓ MMS-TTS

✓ MMS-LID

✓ Omnilingual ASR

✓ WaxalNLP

Benchmarks

ASR (Speech-to-Text)

MALIBA-AI Bambara ASR Benchmark

Model	WER	CER	Source
MALIBA-AI/bambara-asr-v3	45.73	13.45	reported
djelia/asr-v2	47.5	13.56	reported
djelia/asr-v1	48.56	13.0	reported
RobotsMali/soloba-ctc-0.6b-v3	46.76	16.02	reported
RobotsMali/soloni-114m-tdt-ctc-v3-ctc-dec	48.32	14.81	reported

MT (Machine Translation)

MAFAND-MT

Model	spBLEU	chrF++	direction	BLEU	Source
facebook/nllb-200-3.3B	7.7	29.9	fra→bam	—	reported
facebook/nllb-200-3.3B	14.6	37.5	bam→fra	—	reported
masakhane/m2m100_418M_fr_bam_rel_news_ft	—	49.9	fra→bam	24.7	reported
masakhane/m2m100_418M_bam_fr_rel_news_ft	—	49.0	bam→fra	25.8	reported

BOUQuET

Model	direction	chrF++	Source
OMT-LLaMA 8B	eng→bam	?	reported
OMT-LLaMA 8B	bam→eng	?	reported

FLORES+ 10-sample

Model	direction	SpBLEU	ChrF	Source
anthropic/claude-opus-4.6	bam→eng	15.7	43.0	reported
anthropic/claude-opus-4.6	eng→bam	4.1	33.6	reported
baidu/ernie-4.5-300b-a47b	eng→bam	5.9	34.0	reported
baidu/ernie-4.5-300b-a47b	bam→eng	5.9	29.7	reported
google/gemini-2.5-flash	bam→eng	13.7	40.7	reported
google/gemini-2.5-flash	eng→bam	5.6	29.3	reported
google/gemini-3-pro-preview	bam→eng	22.6	46.5	reported
google/gemini-3-pro-preview	eng→bam	5.7	33.1	reported
openai/gpt-5	eng→bam	7.4	34.2	reported
openai/gpt-5	bam→eng	8.1	25.1	reported

GATONES

Model	direction	BLEU	chrF	Source
google/madlad400-3b-mt	bam→eng	11.9	32.8	reported
google/madlad400-3b-mt	eng→bam	6.5	28.5	reported
google/madlad400-7.2b-mt	bam→eng	10.6	31.7	reported
google/madlad400-7.2b-mt	eng→bam	6.1	27.2	reported
google/madlad400-10.7b-mt	bam→eng	12.6	33.9	reported
google/madlad400-10.7b-mt	eng→bam	7.1	30.5	reported
google/madlad400-8b-lm	bam→eng	1.2	14.4	reported
google/madlad400-8b-lm	eng→bam	0	2.1	reported

FLORES-200

Model	direction	BLEU	chrF	Source
google/madlad400-3b-mt	bam→eng	14.1	37.1	reported
google/madlad400-3b-mt	eng→bam	4.9	27.8	reported
google/madlad400-7.2b-mt	bam→eng	12.9	36.3	reported
google/madlad400-7.2b-mt	eng→bam	4.3	26.8	reported
google/madlad400-10.7b-mt	bam→eng	15.5	39.4	reported
google/madlad400-10.7b-mt	eng→bam	5.7	30.1	reported
google/madlad400-8b-lm	bam→eng	2.3	18.8	reported
google/madlad400-8b-lm	eng→bam	0.1	7.8	reported

LLM (Large Language Models)

FLORES-200

Model	Log-perplexity	Source
goldfish-models/bam_latn_full	158.88	reported
MaLA-LM/mala-500-10b-v2	143.14	reported

Winogrande (translated)

Model	Accuracy	Source
GPT-4o	50.2	reported
GPT-4	50.7	reported
meta-llama/Llama-3-70B-Instruct	50.5	reported
BLOOMZ-7b1	49.3	reported

MMLU College Medicine (translated)

Model	Accuracy	Source
GPT-4o	38.2	reported
GPT-4	30.1	reported
meta-llama/Llama-3-70B-Instruct	24.3	reported
BLOOMZ-7b1	26.0	reported

MMLU Clinical Knowledge (translated)

Model	Accuracy	Source
GPT-4o	40.0	reported
GPT-4	34.7	reported
meta-llama/Llama-3-70B-Instruct	32.5	reported
BLOOMZ-7b1	26.8	reported

MMLU Virology (translated)

Model	Accuracy	Source
GPT-4o	32.5	reported
GPT-4	31.9	reported
meta-llama/Llama-3-70B-Instruct	35.5	reported
BLOOMZ-7b1	21.1	reported

Belebele

Model	Accuracy	Source
GPT-4o	38.3	reported
GPT-4	37.9	reported
meta-llama/Llama-3-70B-Instruct	35.1	reported
BLOOMZ-7b1	24.3	reported

SIB-200 10-sample

Model	Classification Acc.	Source
anthropic/claude-3.7-sonnet	100.0	reported
baidu/ernie-4.5-300b-a47b	100.0	reported
openai/gpt-5-mini	100.0	reported

ARC Synthetic 10-sample

Model	ARC Acc.	Source
anthropic/claude-haiku-4.5	90.0	reported
anthropic/claude-opus-4.5	90.0	reported
anthropic/claude-sonnet-4.6	100.0	reported

MGSM Synthetic 10-sample

Model	MGSM Acc.	Source
anthropic/claude-opus-4.5	100.0	reported
anthropic/claude-opus-4.6	100.0	reported
google/gemini-3-pro-preview	80.0	reported

MMLU Synthetic 10-sample

Model	MMLU Acc.	Source
anthropic/claude-opus-4.5	90.0	reported
anthropic/claude-opus-4.6	90.0	reported
google/gemini-2.5-pro	90.0	reported

SAHARA

Model	Lang. Avg.	spBLEU fra→bam	NER F1	Reading comprehension Acc.	Sentiment F1	Source
gemini-3-pro-preview	41.54	11.29	42.11	60.53	52.24	reported
gemini-2.5-pro	33.9	9.56	31.11	47.37	47.57	reported
gpt-5-2025-08-07	27.68	2.69	40.0	23.68	44.36	reported
gpt-5.1-2025-11-13	30.58	2.8	32.56	39.47	47.49	reported
gemini-2.5-flash	23.84	4.31	5.0	39.47	46.6	reported
claude-sonnet-4-5@20250929	23.73	3.58	18.75	39.47	33.12	reported
claude-4-sonnet-20250514	23.82	4.73	16.67	34.21	39.68	reported
gpt-4.1	23.12	1.67	7.41	34.21	49.2	reported
CohereForAI/c4ai-command-a-03-2025	18.59	1.11	0.0	31.58	41.66	reported
meta-llama/Llama-4-Scout-17B-16E-Instruct	24.95	1.7	23.08	31.58	43.42	reported
meta-llama/Llama-3.3-70B-Instruct	21.09	1.02	5.88	31.58	45.89	reported
google/gemma-3-27b-it	17.64	1.15	0.0	26.32	43.09	reported
google/gemma-2-27b-it	19.8	0.82	17.39	28.95	32.05	reported
deepseek-ai/DeepSeek-R1-Distill-Qwen-32B	21.29	0.67	10.53	39.47	34.47	reported
meta-llama/Llama-3.1-70B-Instruct	9.98	0.86	0.0	13.16	25.9	reported
Qwen/Qwen3-30B-Instruct-2507	18.3	0.41	0.0	28.95	43.82	reported
deepseek-ai/DeepSeek-R1-Distill-Llama-70B	15.94	0.71	0.0	28.95	34.1	reported
Tower-Babel/Babel-9B-Chat	14.7	0.62	0.0	26.32	31.85	reported
Tower-Babel/Babel-83B-Chat	10.84	0.79	0.0	23.68	18.88	reported
Qwen/Qwen3-4B-Instruct-2507	19.32	0.48	0.0	34.21	42.58	reported
google/gemma-2-9b-it	18.75	0.43	6.9	18.42	49.24	reported
google/gemma-3-12b-it	12.27	1.06	0.0	10.53	37.49	reported
meta-llama/Llama-3.1-8B-Instruct	14.84	0.55	0.0	18.42	40.4	reported
CohereForAI/aya-23-35B	14.11	0.52	0.0	21.05	34.87	reported
CohereForAI/c4ai-command-r-plus-08-2024	11.47	0.81	0.0	18.42	26.64	reported
CohereForAI/c4ai-command-r7b-12-2024	12.02	0.44	0.0	13.16	34.46	reported
microsoft/Phi-4-mini-instruct	18.39	0.66	0.0	31.58	41.34	reported
meta-llama/Llama-3.2-1B-Instruct	16.25	0.1	0.0	26.32	38.58	reported
microsoft/Phi-3.5-mini-instruct	10.88	0.59	0.0	28.95	13.99	reported
google/gemma-2-2b-it	15.76	0.29	0.0	21.05	41.71	reported
CohereForAI/aya-23-8B	17.01	0.51	0.0	28.95	38.58	reported
google/gemma-3-4b-it	3.49	0.64	0.0	10.53	2.8	reported
meta-llama/Llama-3.2-3B-Instruct	10.34	0.42	0.0	10.53	30.4	reported

📝 Notes

Observations

Survey Focus 1 for Mali (2026 UNICEF WCAR Local Languages Survey)

ASR Models (32)

Model	Size	Downloads	Likes
MahmoudAshraf/mms-300m-1130-forced-aligner	0.3B	4.0M	83
facebook/mms-1b-all	1.0B	2.0M	195
facebook/mms-1b-fl102	1.0B	8.7K	29
UBC-NLP/Simba-S	2B	2.2K	4
facebook/mms-1b-l1107	1.0B	280	12
kalilouisangare/whisper-small-bambara-v2-kis	0.2B	70	1
espnet/xeus	—	60	146
RobotsMali/soloni-114m-tdt-ctc-v0	—	40	3
FarmRadioInternational/bambara-whisper-asr	0.8B	36	0
UBC-NLP/Simba-M	—	33	0
RobotsMali/soloba-ctc-0.6b-v1	—	31	0
RobotsMali/stt-bm-quartznet15x5-v0	—	26	1
RobotsMali/soloni-114m-tdt-ctc-v1	—	23	0
johntsi/ZeroSwot-Medium_asr-cv_en-to-200	0.4B	17	0
johntsi/ZeroSwot-Large_asr-cv_en-to-200	0.4B	16	1
RobotsMali/soloni-114m-tdt-ctc-v3	—	16	0
johntsi/ZeroSwot-Medium_asr-mustc_en-to-200	0.4B	14	0
johntsi/ZeroSwot-Large_asr-mustc_en-to-200	0.4B	12	0
RobotsMali/soloba-ctc-0.6b-v0	—	11	1
RobotsMali/soloba-tdt-0.6b-v0.5	—	11	0

View on HuggingFace: bm: 30 | bam: 2

TTS Models (7)

Model	Size	Downloads	Likes
sudoping01/bambara-tts	36.3M	371	1
ynnov/ekodi-bambara-tts-female	36.3M	59	0
ynnov/ekodi-bambara-tts	36.3M	21	0
rnjema-unima/mms-tts-bam-baseline	36.3M	17	0
MALIBA-AI/bambara-tts	0.5B	11	5
Dapor/Dapson	—	0	4
MALIBA-AI/malian-tts	—	0	0

View on HuggingFace: bm: 6 | bam: 1

MT Models (37)

Model	Size	Downloads	Likes
facebook/nllb-200-distilled-600M	—	1.1M	880
google/madlad400-3b-mt	3B	329.0K	192
facebook/nllb-200-distilled-1.3B	—	147.0K	181
facebook/nllb-200-3.3B	—	81.6K	443
jbochi/madlad400-3b-mt	3B	76.4K	154
facebook/nllb-200-1.3B	—	31.4K	71
Xenova/nllb-200-distilled-600M	—	4.4K	50
mradermacher/OmniTranslate-1.0-GGUF	0.6B	1.1K	3
mradermacher/OmniTranslate-1.1-GGUF	0.6B	1.1K	2
JustFrederik/nllb-200-distilled-600M-ct2-int8	—	925	1
facebook/nllb-moe-54b	—	557	123
MihaiPopa-1/OmniTranslate-1.1	0.6B	393	0
MihaiPopa-1/OmniTranslate-1.0	0.6B	374	0
JustFrederik/nllb-200-1.3B-ct2-int8	—	200	0
JustFrederik/nllb-200-distilled-600M-ct2-float16	—	157	0
winstxnhdw/nllb-200-distilled-1.3B-ct2-int8	—	141	6
JustFrederik/nllb-200-distilled-1.3B-ct2-int8	—	116	2
JustFrederik/nllb-200-distilled-600M-ct2	—	65	2
JustFrederik/nllb-200-3.3B-ct2-float16	—	52	3
michaelfeil/ct2fast-nllb-200-distilled-1.3B	—	47	1

View on HuggingFace: bm: 30 | bam: 7

LLM Models (36)

Model	Size	Downloads	Likes
bigscience/bloomz-560m	0.6B	1.0M	137
bigscience/bloom-560m	0.6B	164.0K	371
bigscience/bloom-1b7	2B	49.4K	122
bigscience/bloom-3b	3B	14.1K	94
bigscience/bloomz-3b	3B	13.3K	81
bigscience/bloom-7b1	7B	10.4K	202
bigscience/bloom-1b1	1B	7.0K	66
bigscience/bloom	176B	5.9K	5.0K
bigscience/bloomz-7b1	7B	4.2K	147
bigscience/bloomz-7b1-mt	7B	4.1K	144
bigscience/bloom-7b1-intermediate	7B	3.7K	0
bigscience/bloomz-7b1-p3	7B	3.5K	6
bigscience/bloomz	176B	3.5K	513
bigscience/bloom-1b1-intermediate	1B	3.4K	0
bigscience/bloomz-mt	176B	3.3K	32
bigscience/bloomz-1b1	1B	2.9K	33
bigscience/bloomz-1b7	2B	2.8K	23
UBC-NLP/toucan-base	0.6B	2.5K	5
bigscience/bloom-1b7-intermediate	2B	2.4K	0
bigscience/bloom-560m-intermediate	0.6B	2.2K	1

View on HuggingFace: bm: 30 | bam: 6

Datasets

ASR Datasets (28)

Dataset	Rows	Downloads	Likes
sarulab-speech/yodas2_sidon	1.5M	6.0K	60
BolajiJsAI/WaxalNLP	2.6M	2.5K	1
suryatmodulus/yodas2_sidon	—	1.1K	0
WueNLP/belebele-fleurs	65.1K	799	9
RobotsMali/kunkado	119.0K	469	4
RobotsMali/afvoices	875.0K	400	5
RobotsMali/bam-asr-early	79.8K	296	6
WueNLP/sib-fleurs	94.4K	260	11
uvci/koumankan4dyula	10.9K	234	6
soynade-research/Bambara-Speech-Translation-Data	221.0K	207	1
YuRiVeRTi/V1Q	186.0K	201	3
OumarDicko/Bambara_AudioSynthetique_42K_V3	42.0K	79	2
RobotsMali/jeli-asr	67.3K	47	4
RobotsMali/Bam_ASR_Eval_500	500	31	1
RobotsMali/transcription-scorer	1.2K	23	1
MALIBA-AI/bambara-asr-benchmark	500	23	0
sil-ai/bloom-speech	—	16	25
french-datasets/espnet_yodas2	—	16	0
binaryMao/Kunkado_Maya_V2	119.0K	14	0
sudoping01/bam-asr-benchmark	500	13	0

View on HuggingFace: bm: 24 | bam: 4

TTS Datasets (17)

Dataset	Rows	Downloads	Likes
sarulab-speech/yodas2_sidon	1.5M	6.0K	60
BolajiJsAI/WaxalNLP	2.6M	2.5K	1
suryatmodulus/yodas2_sidon	—	1.1K	0
WueNLP/belebele-fleurs	65.1K	799	9
RobotsMali/bam-asr-early	79.8K	296	6
WueNLP/sib-fleurs	94.4K	260	11
YuRiVeRTi/V1Q	186.0K	201	3
RobotsMali/jeli-asr	67.3K	47	4
Kabeer9997/Mix.AI-Data	—	19	0
sil-ai/bloom-speech	—	16	25
binaryMao/Kunkado_Maya_V2	119.0K	14	0
lenoxaugo254/LenoxAI	—	10	1
Menoviar28/menov31	—	9	0
dijihax/Dataset	—	8	2
oza75/bambara-tts	275.0K	7	5
djelia/bambara-audio	633.0K	5	1

View on HuggingFace: bm: 15 | bam: 2

Translation Datasets (46)

Dataset	Rows	Downloads	Likes
HuggingFaceFW/finetranslations	—	75.8K	283
facebook/flores	—	43.4K	97
openlanguagedata/flores_plus	893.0K	15.8K	124
gtak1/panlex-meanings	64.1M	6.3K	0
google/smol	798.0K	2.8K	94
Muennighoff/flores200	—	1.6K	20
facebook/bouquet	5.6M	1.6K	33
aiana94/polynews-parallel	5.4M	1.4K	13
ayymen/Pontoon-Translations	3.6M	1.1K	18
HuggingFaceFW/finetranslations-edu	109.0M	1.1K	27
Helsinki-NLP/tatoeba	—	866	56
openlanguagedata/oldi_seed	564.0K	772	11
masakhane/mafand	143.0K	463	15
ayymen/Weblate-Translations	11.7M	383	18
sbmaruf/forai_ml_masakhane_mafand	136.0K	365	2
mteb/FloresBitextMining	2.0K	303	1
RobotsMali/bam-asr-early	79.8K	296	6
visheratin/laion-coco-nllb	894.0K	255	44
cointegrated/panlex-meanings	157.0M	255	13
uvci/koumankan4dyula	10.9K	234	6

View on HuggingFace: bm: 30 | bam: 16