Skip to main content
HomeAPI DocsPricing
AI Models by kiwi

Current API Models, Clearly Tiered

Free API keys can call auto as the recommended default endpoint model and hrllm directly at 40 requests per hour. PRO unlocks direct access to the rest of the new lineup, while legacy Kiwi models remain visible as deprecated compatibility entries.

Legacy Kiwi aliases are deprecated and marked with EOL: 09.03.2026. They stay visible for lineage and transition planning, but the current public lineup centers on AUTO (auto), hrLLM (hrllm), and the new direct PRO models.

Recommended Default Endpoint Model

FreeRecommended
AUTO
Default public endpoint model. The system automatically selects the best current model lane for the request.

AUTO is the default public route for English and general-purpose use. It keeps integrations simple, removes model-choice friction in the dashboard, and lets the platform choose the best current lane automatically.

Best default choice when you want dependable English output without manually picking a direct model.

Matches the recommended public API path shown in docs, examples, and dashboard defaults.

Lets the platform route requests to the best current lane automatically.

Public API access

Public model ID: auto

Free API keys: 40 requests/hour

Recommended for English chat, quick starts, dashboard defaults, and teams that do not want to manage direct model selection.

Use hrllm directly when the workflow must stay Croatian-only.

Open AUTO page

Current API Lineup

These are the current public-facing models in the lineup. AUTO is the recommended default route, hrLLM remains the free direct Croatian model, and the other direct models are available as PRO.

FreeRecommended
AUTO
auto

api.llm.kiwi

Recommended default endpoint model for English and general-purpose use across the API and dashboard.

Best used for: English chat, general-purpose prompts, quick integrations, dashboard defaults, and teams that want the best current lane selected automatically.

Based on: auto

Open model page
PRO
GPT-OSS-120B
gpt-oss-120b
OSS

api.llm.kiwi

Large-scale open-source Pro model for comprehensive analysis and extensive reasoning tasks.

Best used for: Large-scale analysis, comprehensive reasoning, extensive technical documentation, and complex multi-domain tasks.

Based on: GPT-OSS-120B

Open model page
Free
hrLLM
hrllm

api.llm.kiwi

Croatian-first model for writing and answering in grammatically correct Croatian.

Best used for: Croatian customer support, formal business writing, public-sector communication, and education content.

Based on: hrllm

Open model page
PRO
Kimik2
kimik2

api.llm.kiwi

Advanced Pro model for complex reasoning, technical analysis, and sophisticated problem solving.

Best used for: Complex technical tasks, advanced reasoning, multi-step analysis, and sophisticated problem solving.

Based on: kimik2

Open model page
PRO
Qwen3
qwen3

api.llm.kiwi

Compact Pro model for quick reasoning, drafting, and lightweight production tasks.

Best used for: Fast general chat, structured drafting, lightweight copilots, and low-latency automations.

Based on: Qwen3-1.4B

Open model page

Deprecated Models

Deprecated models remain listed for continuity, migration, and provider lineage. They are intentionally greyed out and clearly marked with their EOL date.

Access and Usage Limits

These are the model-access highlights users need most often. The complete reference stays in the docs.

Public API chat
POST https://api.llm.kiwi/v1/chat/completions

Free: 40 requests/hour for AUTO and hrLLM

PRO unlocks direct access to the new advanced models with higher sustained throughput.

Public model catalog
GET https://api.llm.kiwi/v1/models

192 requests/minute per IP

Cache-friendly endpoint for model discovery and compatibility metadata.

AI Tools Web
POST /api/internal/chat

36 requests/minute per signed-in user + IP

Cookie-authenticated dashboard execution for configured client tools.

Tools proxy
POST /api/tools/proxy/chat

Account-key and tool-specific limits

Server-side tool execution with request logging and tool_id targeting.