Mohloli oa Sentiment
Ho sebelisa khoebo
579 models
Ha ho na letšoao la metsi
Ha ho hlokahale ho ngolisa
Mofuta:
+ GPT-5, Claude, Gemini
Litšoantšo
-
-
Bophara
-
ho tloha ho -100 ho ea ho +100
Boikutlo
-
-
Litlhaku
Likhetho tse tsoetseng pele
Phello
Tokens e tlase.
Fumana Token e eketsehileng
❤️ U rata Free.ai? Reka ho ba lelapa la hao!
Register ho fumana sehokela sa ho u joetsa le ho fumana li-token tse 25 000 ka motsoalle.
U batla ho feta?
E-ba le tokelo ea ho ingolisa: 30,000 tokens/day
Ho ngolisa
Ho sebetsana le kopo ea hau...
Ho hlahloba maikutlo a mongolo ka AI e mahala. Ho bona li-tone tse ntle, tse mpe le tse sa tšoaneng.
Ho sebelisa Mohloli oa Sentiment
1
Tobetsa ho kenya
Tlatsa mongolo, kenya faele, kapa hlalosa seo u se batlang. Ha ho hlokahale ak'haonte.
2
Tobetsa ho theha
AI ea rona e sebetsana le kopo ea hau ka metsotsoana e seng mekae ka ho sebelisa li-models tse ntlehali tsa open-source.
3
Tlosa & arolelana
Kopitsa, kenya kapa arolelana sephetho sa hau. Haholo-holo bakeng sa ho sebelisana le batho ba bang le ho rekisa.
Senya sesebelisoa sena ka API
Automatize ena sesebelisoa ho tloha ho hao code. OpenAI-ka lumellanang REST endpoint, Bearer-token auth, ha ho hlokahale SDK e eketsehileng. Token theko e lumellana le web interface.
curl -X POST https://api.free.ai/v1/chat/ \
-H "Authorization: Bearer sk-free-..." \
-H "Content-Type: application/json" \
-d '{"model": "qwen7b", "messages": [{"role": "user", "content": "Summarize this: ..."}]}'Li-tools tsa AI tse amanang le mahala
Mohloli oa Sentiment - FAQ
Lintlha tse tharo, ho latela mokhoa o u khethileng: (1) polarity e felletseng - -100 ho +100 ka sekhahla le tšepo; (2) li-emotions tse 6-dimensional - thabo, khotso, ho tšoenyeha, ho tšoenyeha; (3) ho latela litšobotsi - ho tlosa litšobotsi tse bitsoang kapa li-entities le ho etsa sekhahla ka bobeli ka ho khetheha; (4) sarcasm / subtext - ho fetola polarity ea lefatše ka irony e fumanoeng; (5) ka-senang - ho tag lentsoe le leng leng ka ho khetheha. Khetha mokhoa o lumellanang le ketsahalo ea hau ea ho sebelisa.
Ehlile - 200-lentsoe le tloaelehileng la tlhaloso e sebetsa ka ho sa feleng Qwen 3 30B moralo ka ~ 200 tokens, ka khotso ka hare 6,000 Anonymous kapa 30,000 kopanetsoeng-ho fihlela letsatsi le letsatsi pool. Forensic-bophahamo bophahamo botlhahlobo ka lingoloa tse telele theko ~ 600 tokens. Ha ho na ho ngola ho hlokahala ho leka tse ling.
Ho bapisoa ka ho feletseng (~ 85% tšebelisano'moho le barekisi ba batho ka mongolo o hlakileng), ka morao ho Google ka ho nepahala ha ho hlophisoa ha lihlooho, ka pele ho tse ngata ka sarcasm hobane re fa moralo oa ho bonts'a ho bonts'a sarcasm. Barekisi ba Enterprise ba lefa $ 1-4 ka 1,000 units.
Yeah - fetola ho "Sarcasm + subtext" mode. The system prompt explicitly instructs the model to flip surface polarity when it detects irony ("Oh great, another delay" scores negative despite positive-word presence). Accuracy on apparent sarcasm is ~80%; subtle passive-aggression is harder for any model. The "Mixed" verdict + low-confidence score is your flag to read the text yourself.
Instead of one verdict for the whole text, you get separate scores for each entity or feature mentioned. Example: "The hotel was beautiful but the breakfast was cold and the wifi was terrible" yields {hotel: +80, breakfast: -50, wifi: -70}. Ideal for product reviews, restaurant feedback, and NPS comments where customers mix praise and complaints in one paragraph.
99 languages. The six emotion dimensions come from Ekman basic-emotion research and translate universally. Accuracy is highest on English, Spanish, French, Portuguese, German, Chinese, Japanese; lower-resource languages work but may produce slightly lower confidence scores.
Ehlile - e 'ngoe ea liketsahalo tsa eona tse phahameng ka ho fetisisa. Ea litekete tse tsoang ka mokhoa oa "Support ticket" domain; litekete tse nang le li-scores tse tlase ho -50 polarity le ho phahama ha ho phahama ho ea ho tšehetso ea hau ea senior, litekete tsa positive-verdict li tsamaea ho sebeletsa. Eba JSON ka ho tobetsa ho kenya le ho kenya ho Zendesk / Freshdesk automations.
Ea - POST ho /v1/chat/ le lengolo-tsoibila le ts'oanang le lengolo-tsoibila lena le thehiloe (bona sehlooho bakeng sa lengolo-tsoibila le nepahetseng). E khutlisa JSON e hlophisitsoeng. E loketse dashboards tse nang le limilione tsa litlhahlobo ka letsatsi. Bearer auth, liphelelo tsa khoeli. Docs ho /api/.
Boikutlo bo khotsofatsang bo bonts'a boikutlo bo khotsofatsang ba sehlooho, eseng boikutlo ba hau. Li-input tse khuts'oane (< 20 mantsoe), lingoloa tse sa utloahaleng, le litaba tse nang le ho tšoana ha li-polarities li fumana lipalo tse tlase. Boikutlo bo khotsofatsang bo tlase bo ke ke ba le boikutlo bo sa khotsofatsang - ke "Ke sa tsebe". E-ba le lipalo life kapa life tse tlase ho 50 e le hore li hloka ho hlahlojoa ke motho.
UI e sebetsa ka mongolo o le mong ka nako. Bakeng sa palo e kholo, ho kenya JSON + CSV ho u lumella ho hokahanya API ho Python kapa Node script e iterates dataset ea hau. Lipheo: 20,000 li-characters ka ho ngola, 100 li-calls / hora ka IP bakeng sa basebelisi ba sa tsejoeng, 1,000 / hora bakeng sa ak'haonte tse ngotsoeng. Lipheo tsa Enterprise rate li fumaneha - ikopanye le rona.
MonkeyLearn e ne e fumaneha'me e sa tsoa emisa. Lexalytics le Repustate li lefa $ 500-5,000 / mo bakeng sa ho fumana maikutlo a amanang le maikutlo a amanang le maikutlo. Mochini oa rona o mahala bakeng sa li-workflows tse ngata; li-models tsa premium (GPT-5, Claude Sonnet) tse fumanehang ka ho ngola bakeng sa liketsahalo tsa ho sebelisa liketsahalo tse phahameng joalo ka li-transcripts tsa ho ngola.
Lingoloa li sebetsana ka mohopolo ka GPU ea rona, ha li sa buloa disk bakeng sa basebelisi ba sa tsejoeng. Basebelisi ba nang le ak'haonte ba bona lingoloa ka dashboard ea bona ka matsatsi a 7. Re ke ke ra arolelana lingoloa le batho ba boraro kapa ra li sebelisa bakeng sa ho ithuta. Bakeng sa ho latela zero-log sebelisa sebaka sa hau sa ho qetela sa API-key-gated - ikopanya le rona bakeng sa SLAs tsa k'hamphani.
E-ba le tokelo ea ho ingolisa: 30,000 tokens/day
E etsa akhaontoHa ho hlokahale karete ea mokitlane
U tla lekola joang sena?