Boshlash

Moslik

Ailoot qaysi API formatlarini tushunishi, har birida qaysi modellar mavjudligi va ulardan qaysi mijoz dasturlari foydalanishi.

To‘liq qo‘llanma

Ailoot uchta API’ning ommaviy kontraktlarini takrorlaydi. Dastur OpenAI, Anthropic yoki Gemini bilan ishlay olsa va o‘z manzilingizni ko‘rsatishga imkon bersa, u Ailoot bilan ham ishlaydi — manzil va kalitni almashtirish kifoya.

Barcha yo‘llar quyidagi manzilga nisbatan berilgan: https://api.ailoot.ai. «SDK uchun manzil» ustunida bazaviy manzil sifatida nimani ko‘rsatish kerakligi berilgan: https://api.ailoot.ai/v1 (/v1 bilan) yoki https://api.ailoot.ai (/v1’siz), chunki SDK’lar yo‘lning qolgan qismini o‘zi qo‘shadi.

Format	Yo‘l	SDK uchun manzil	ID
OpenAI Chat Completions	`POST /v1/chat/completions`	/v1 bilan	139
OpenAI Responses	`POST /v1/responses`	/v1 bilan	106
Anthropic Messages	`POST /v1/messages`	/v1’siz	150
Gemini	`POST /v1beta/models/{m}:generateContent` `POST /v1beta/models/{m}:streamGenerateContent`	/v1beta’siz	29
OpenAI Embeddings	`POST /v1/embeddings`	/v1 bilan	6
OpenAI Images	`POST /v1/images/generations` `POST /v1/images/edits`	/v1 bilan	30
OpenAI Audio	`POST /v1/audio/speech` `POST /v1/audio/transcriptions`	/v1 bilan	20
Video va musiqa	`POST /v1/videos/generations` `POST /v1/music/generations`	/v1 bilan	18
Qarorlar (Jev)	`POST /v1/systemone`	SDK’siz, to‘liq manzil	1
Modellar ro‘yxati (OpenAI, Anthropic)	`GET /v1/models`	/v1 bilan	barchasi
Modellar ro‘yxati (Gemini)	`GET /v1beta/models`	/v1beta’siz	barchasi

Katalogda 1256 ta model kartochkasi va hujjatlashtirilgan 2692 ta metod mavjud. Muqobil chaqiruv identifikatorlari yoki taxalluslar mavjud bo‘lsa, tanlangan metod ostida ko‘rsatiladi.

Qaysi model qaysi formatda

Matn modellarining aksariyati — Claude, GPT, DeepSeek, Qwen, Kimi, GLM, MiniMax va Grok — bir vaqtning o‘zida uchta formatda mavjud: OpenAI Chat, OpenAI Responses va Anthropic Messages. Gemini modellari — Anthropic Messages va o‘z Gemini formatida. NVIDIA modellari asosan OpenAI Chat formatida.

OpenAI Chat Completions

claude-sonnet-5-5, gpt-6.1-sol, gemini-3.6-flash, grok-4.7, deepseek-v4-pro, qwen3.8-max, k3, glm-5.3, MiniMax-M3, nvidia/nemotron-3-ultra-550b-a55b, meta/muse-glimmer-30b, poolside/laguna-xs-2.1, claude-opus-5-5, gpt-6-astra va yana 51

OpenAI Responses

claude-sonnet-5-5, gpt-6.1-sol, grok-4.7, deepseek-v4-pro, qwen3.8-max, k3, glm-5.3, MiniMax-M3, claude-opus-5-5, gpt-6-astra, grok-4.7-build-fast, deepseek-v4.1-flash, qwen3.8-flash, kimi-for-coding va yana 36

Anthropic Messages

claude-sonnet-5-5, gpt-6.1-sol, gemini-3.8-flash, grok-4.7, deepseek-v4-pro, qwen3.8-max, k3, glm-5.3, MiniMax-M3, nvidia/nemotron-3-ultra-550b-a55b, meta/muse-glimmer-30b, poolside/laguna-xs-2.1, claude-opus-5-5, gpt-6-astra va yana 71

Gemini

gemini-3.8-flash, gemini-3.1-pro, gemini-pro-agent, gemini-3.8-flash-tiered, gemini-3.8-flash-medium, gemini-3.8-flash-low, gemini-3.8-flash-high, gemini-3.7-flash, gemini-3.7-flash-tiered, gemini-3.7-flash-medium, gemini-3.7-flash-low, gemini-3.7-flash-high, gemini-3.6-flash, gemini-3.6-flash-tiered va yana 13

Narxlar va filtrlar bilan to‘liq ro‘yxat — modellar katalogida.

Qaysi mijoz dasturi qaysi formatdan foydalanadi

Quyida integratsiyalar katalogimizdagi ilovalar foydalanadigan formatiga qarab guruhlangan. Ko‘pchiligi bir nechta formatni qo‘llab-quvvatlaydi.

Formatlarda nimalar qo‘llab-quvvatlanadi

Imkoniyat	Qanday ishlaydi
Oqimli javob	Barcha matn formatlarida SSE formati: `stream: true` yoki `:streamGenerateContent?alt=sse`. Pullik matn narx hisoblangandan keyin so‘zma-so‘z emas, bitta blokda keladi — batafsil
Vositalar (tools, function calling)	Ha, model ularni qo‘llab-quvvatlasa. Vositalar tavsifi tanlangan API formatida beriladi. Gemini matn so‘rovlarida googleSearch va urlContext ham qabul qilinadi — alohida yoki birga, ammo ayni so‘rovda functionDeclarations bilan birga emas. Nativ rasm chiqishi uchun faqat mijoz taqdim etgan deklaratsiyalar qo‘llab-quvvatlanadi: `functionDeclarations`
Kirishdagi rasmlar	Ha, tasvirlarni ko‘ra oladigan modellarda. Rasmlar narxining taxminiy bahosi zaxiraga kiradi. Gemini formatida — PNG, JPEG yoki WebP `inlineData` maydonida (foydalanuvchi xabarida)
Tuzilmali javob (JSON)	`response_format` OpenAI’da va uning Anthropic’dagi muqobillari — model qo‘llab-quvvatlasa. Gemini formatida `responseMimeType` generationConfig ichida text/plain yoki application/json qiymatini qabul qiladi; aniq model qo‘llab-quvvatlaydigan responseSchema yoki responseJsonSchema’dan foydalaning. Imkoniyatlar va parametrlarni birga qo‘llash variantlari modelga bog‘liq
Mulohaza yuritish	`reasoning_effort` / `reasoning` OpenAI’da, `thinking` Anthropic’da, `thinkingConfig` Gemini’da — mulohaza yurituvchi modellar uchun
Tokenlarni hisoblash	Tokenlarni alohida hisoblash (`/v1/messages/count_tokens`) — alohida metod; uning mavjudligi tanlangan model va kalit uchun tekshiriladi. Messages qo‘llab-quvvatlanishi tokenlarni hisoblash ham qo‘llab-quvvatlanishini o‘z-o‘zidan tasdiqlamaydi

Savollaringiz qoldimi? Request ID va tavsifni yuboring — tushunishga yordam beramiz.Yordam xizmatiga yozish