Nhazigharị

Ihe iri ise, n'ihe banyere móòdù ị họọrọ: (1) n'ogbe polarity - a -100 ruo +100 n'ọnụọgụgụ na nghọta; (2) 6-dimensional emotions - obi ụtọ, ọdachi, nchegbu, ọdachi, ihe a na-agaghị ekwe omume, n'anụ ọkụ n'obi; (3) aspektị-n'okpuru - na-apụta aha ihenhọrọ mọọbụ ihenhọrọ na n'ọnụọgụgụ nke ọbụla; (4) sarcasm / subtext - na-atụgharị n'elu polarity na ihenhọrọ a na-ahụkarị; (5) per-sentence - tagụọ okwu ọbụla n'otu. Họrọ módù nke na-adabere na ọnọdụ ịrụ ọrụ gị.

Ee - a 200-okwu na-akọwapụta na-agba ọsọ Qwen 3 30B diff model na ~ 200 tokens, n'ụzọ dị mfe n'ime 2,500 anonymous ma ọ bụ 10,000 signed-up daily pool. Forensic-depth aspect analysis on longer texts costs ~ 600 tokens. No sign-up required to try a few.

Nwere ike ịtụle na ntụgharị zuru oke (~ 85% nkwekọrịta na ndị na-enyocha mmadụ na ntinye ederede dị mfe), n'azụ Google na nghọta nke ihe nchọgharị, n'ihu ọtụtụ na sarcasm n'ihi na anyị na-enye ụdị ihe nchọgharị na-egosipụta ihe nchọgharị. Ndị na-ere ahịa Enterprise na-akwụ ụgwọ $ 1-4 kwa 1,000 units. Advantage anyị: emotion radar + aspect-based output in one call, free, no API key to provision.

Ya - gbanwee na "Sarcasm + subtext" mode. Sistemụ nkuzi na-agwa móòdù ka ọ gbanwee n'elu ala mgbe ọ na-ahụta ihe na-atọ ụtọ ("Oh great, ọzọ delay" na-emeri n'ihi na okwu na-atọ ụtọ nọ). Nhazi na sarcasm dị n'anya bụ ~80%; n'ihi na ọ dị mfe ịrụ ọrụ na-esiri ike maka módù ọbụla. "Mixed" verdict + ala-n'aka na-emeri bụ akara gị ka ịgụ ngwe ahụ onwe gị.

N'ebe ahụ ka ị na-ewere n'aka otu n'ime ngwe niile, ị ga-enweta ihenhọrọ dị iche iche maka ihenhọrọ ọbụla ma ọ bụ ihenhọrọ akọwapụtara. Ụdị: "Obi hotelụ ahụ dị mma mana nri ụtụtụ bụ ọkụ ma ọ bụ wifi bụ ihe ọjọọ" na-enye {hotelụ: +80, nri ụtụtụ: -50, wifi: -70}. Ọ dị mma maka nyocha ngwaahịa, nzaghachi ụlọ oriri na ọṅụṅụ, na nkọwa NPS ebe ndị ahịa na-ejikọta ịtụkwasị obi na ịtụkwasị obi n'otu paragraf.

99 asụsụ. The atọ emotions dimensions na-abịa site Ekman basic-emotions nnyocha na tọghata universally. Accuracy bụ elu na English, Spanish, French, Portuguese, German, Chinese, Japanese; ala-ọnụọgụgụ ọrụ ma nwere ike na-emepụta a bit ala confidence scores.

Ee - nke ahụ bụ otu n'ime ihe ndị kachasị mma. Rụọ ọrụ tiketi na-abịa site na "Tiketi nkwado" mode; tiketi na-egosipụta n'okpuru -50 polarity na-arị elu-dimension gaa na ntọala nkwado gị, tiketi na-egosipụta ụzọ iji hụ onwe gị. Ekpughe JSON site na bọtịnụ nbudata na waya na Zendesk / Freshdesk automations.

Ya - POST ka /v1/chat/ na sistem n'otu n'otu na-akpụzi ihuakwụkwọ a (lee isi mmalite maka n'otu n'otu n'otu). Na-eziga JSON nke ejiri arụpụta. Ọ dị mma maka dashboards nke na-echekwa nde ndị nlele n'ụbọchị. Ogo onye na-eweta, n'ọnwa. Doktọ na /api/.

N'aka nke ọzọ, n'aka nke ọzọ, ọ bụghị n'aka gị. Inputs dị n'ime (<20 okwu), ngwe na-akọwaghị, nakwa ihenhọrọ nke na-atọ-ọdịiche-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-n'etiti-

UI na-agbagharị ngwe otu n'otu oge. Maka nnukwu, ibudata JSON+CSV na-enye gị ohere ịkpọgharịa API n'ime Python mọọbụ Node Script nke na-ewere dataset gị. Ogo: 20,000 akara n'otu oku, 100 oku/ọnụọgụgụ n'otu IP maka ndị ọrụ anọnimị, 1,000/ọnụọgụgụ maka akaụntụ signed-up. Ogo ọnụọgụgụ ụlọ ọrụ dị - kpọtụrụ anyị.

MonkeyLearn bụ a na-akpali na-akwụsị. Lexalytics na Repustate ụgwọ $ 500-5,000 / mo maka ekwivalent aspect-based + emotions detection. Anyị tool bụ free maka ọtụtụ workflows; premium models (GPT-5, Claude Sonnet) dị na-akpọ maka ndị kasị elu-stakes iji ihe ndị dị ka ego-akpọ transcripts.

A na-enyocha ngwe n'ime n'ime n'ime GPU anyị, ọ bụghị na-echekwa na diski maka ndị ọrụ na-enweghị aha. Ndị ọrụ na-ejikọta na-ahụ akụkọ na dashboard ha maka ụbọchị 7. Anyị anaghị etinye ngwe na ndị ọzọ ma ọ bụ jiri ya maka nkuzi. Maka nlekọta nke zero-log jiri gị onwe gị API-key-gated endpoint - kpọtụrụ anyị maka SLAs nke ụlọ ọrụ.