Text Generation

Text models power chat, reasoning, tool use, and structured output. You call any text model through one of two API formats — OpenAI or Anthropic — and the router resolves the best backend and translates as needed.

OpenAI format

Endpoint	Use it when…
Chat Completions — POST /v1/chat/completions	You use the OpenAI SDK, LangChain, LiteLLM, or any OpenAI-compatible tool.
Responses — POST /v1/responses	You target the OpenAI Responses API.

Anthropic format

Endpoint	Use it when…
Messages — POST /v1/messages	You use the Anthropic SDK or Claude-native tooling.

Both formats support the same core capabilities — see Features for tool calling, reasoning, structured outputs, prompt caching, and multimodal input.

Quick example

Python (OpenAI SDK)cURL

from openai import OpenAI

client = OpenAI(base_url="https://llm.siraya.ai/v1", api_key="<API_KEY>")
resp = client.chat.completions.create(
    model="claude-opus-4.8",
    messages=[{"role": "user", "content": "Explain routing in one sentence."}],
)
print(resp.choices[0].message.content)

curl https://llm.siraya.ai/v1/chat/completions \
  -H "Authorization: Bearer <API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-opus-4.8","messages":[{"role":"user","content":"Hello!"}]}'

Model families

Browse text models by family — each page lists its members and capabilities. For how parameters behave, see Parameter Support by Model and Per-Model Details.

Family	Highlights
Claude	Long context, strong reasoning, extended thinking.
GPT	Broad capability, reasoning_effort control, coding variants.
Gemini	Fast multimodal models with large context.
Qwen	Versatile general-purpose models.
GLM	General-purpose chat and reasoning.
DeepSeek	Efficient reasoning-oriented models.
Grok	Fast conversational models.
MiniMax	General-purpose chat models.
Kimi	Long-context chat models.
Seed	General-purpose chat models.