/v1/models
接入前查询当前可调用模型 ID。ChatGPT 模型 ID 以此接口返回为准。
文档
API 接入文档
Try AI API 统一网关的接入指南：可用模型、快速开始、调用入口、代码示例、SDK 配置与常见问题。
文本调用 · cURL
curl https://tryaiapi.com/v1/chat/completions \
-H "Authorization: Bearer $TRYAIAPI_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "claude-sonnet-4-6",
"messages": [
{"role": "user", "content": "Hello"}
]
}'
视觉模型任务 · cURL
curl https://tryaiapi.com/videos/v1/videos/generations \
-H "Authorization: Bearer $TRYAIAPI_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "doubao-seedance-2.0",
"prompt": "A cinematic product shot",
"duration": 5,
"resolution": "480p",
"aspect_ratio": "16:9"
}'
Models
文本模型与视觉模型，一个 Key 接入
统一查看 Claude、ChatGPT 与视觉模型的调用入口、模型 ID 获取方式和状态；完整目录见 模型页面。
接口定义见 /docs/openapi.json；模型详情以实时 /v1/models 与 模型市场 为准。
curl · List models
curl https://tryaiapi.com/v1/models \
-H "Authorization: Bearer $TRYAIAPI_KEY"
Quick Start
三步接入
STEP 02
选择调用入口
文本模型替换 SDK base URL；视觉模型使用 /videos/v1/videos/generations 提交任务。
STEP 03
查看结果与用量
文本请求实时返回；视觉模型按任务轮询，完成后可下载结果，并进入使用记录汇总。
API Endpoints
调用入口
/v1/chat/completions
OpenAI 兼容入口。完成 base URL 替换后，OpenAI SDK、Cursor、Codex CLI、LangChain 可走此入口。
/v1/messages
Anthropic Messages 原生入口。Claude Code 与 Anthropic SDK 可直接切换到 Try AI API base URL。
/videos/v1/videos/generations
视觉模型异步任务入口。提交 prompt、时长、清晰度和画幅后返回 job_id 与 poll_url。
/videos/v1/videos/generations/{task_id}
轮询视觉模型任务状态，查看完成结果、失败原因和视频下载地址。
Code Examples · Advanced
进阶场景示例
文本高级能力与视觉模型异步任务示例，每个场景给出 cURL / Python / JavaScript 三种写法。
cURL · 流式响应（SSE）
curl -N https://tryaiapi.com/v1/chat/completions \
-H "Authorization: Bearer $TRYAIAPI_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "claude-sonnet-4-6",
"messages": [{"role": "user", "content": "写一首关于云的短诗"}],
"stream": true
}'
Python · 流式响应
from openai import OpenAI
client = OpenAI(base_url="https://tryaiapi.com/v1", api_key="$TRYAIAPI_KEY")
stream = client.chat.completions.create(
model="claude-sonnet-4-6",
messages=[{"role": "user", "content": "写一首关于云的短诗"}],
stream=True,
)
for chunk in stream:
print(chunk.choices[0].delta.content or "", end="")
JavaScript · 流式响应
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://tryaiapi.com/v1",
apiKey: process.env.TRYAIAPI_KEY,
});
const stream = await client.chat.completions.create({
model: "claude-sonnet-4-6",
messages: [{ role: "user", content: "写一首关于云的短诗" }],
stream: true,
});
for await (const chunk of stream) {
process.stdout.write(chunk.choices[0].delta.content ?? "");
}
cURL · 视觉模型异步任务
# 1. Submit job
curl https://tryaiapi.com/videos/v1/videos/generations \
-H "Authorization: Bearer $TRYAIAPI_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "doubao-seedance-2.0",
"prompt": "A cinematic product shot on a clean studio desk",
"duration": 5,
"resolution": "480p",
"aspect_ratio": "16:9"
}'
# 2. Poll result
curl https://tryaiapi.com/videos/v1/videos/generations/$JOB_ID \
-H "Authorization: Bearer $TRYAIAPI_KEY"
Python · 视觉模型异步任务
import os
import time
import requests
headers = {
"Authorization": f"Bearer {os.environ['TRYAIAPI_KEY']}",
"Content-Type": "application/json",
}
job = requests.post(
"https://tryaiapi.com/videos/v1/videos/generations",
headers=headers,
json={
"model": "doubao-seedance-2.0",
"prompt": "A cinematic product shot on a clean studio desk",
"duration": 5,
"resolution": "480p",
"aspect_ratio": "16:9",
},
)
job.raise_for_status()
poll_url = job.json()["poll_url"]
while True:
result = requests.get(poll_url, headers=headers)
result.raise_for_status()
data = result.json()
if data["status"] in {"completed", "failed", "cancelled", "timeout"}:
print(data)
break
time.sleep(5)
JavaScript · 视觉模型异步任务
const headers = {
Authorization: `Bearer ${process.env.TRYAIAPI_KEY}`,
"Content-Type": "application/json",
};
const submit = await fetch("https://tryaiapi.com/videos/v1/videos/generations", {
method: "POST",
headers,
body: JSON.stringify({
model: "doubao-seedance-2.0",
prompt: "A cinematic product shot on a clean studio desk",
duration: 5,
resolution: "480p",
aspect_ratio: "16:9",
}),
});
const job = await submit.json();
while (true) {
const res = await fetch(job.poll_url, { headers });
const result = await res.json();
if (["completed", "failed", "cancelled", "timeout"].includes(result.status)) {
console.log(result);
break;
}
await new Promise((resolve) => setTimeout(resolve, 5000));
}
视频任务参数参考
| 参数 | 类型 | 说明 |
|---|---|---|
model |
string | doubao-seedance-2.0 或 doubao-seedance-2.0-fast |
prompt |
string | 文本描述，必填 |
duration |
integer | 时长（秒），范围 4–15 |
resolution |
string | 480p / 720p / 1080p（fast 不支持 1080p） |
image_urls |
array ≤9 | 参考图片。每项可以是 URL 字符串（默认角色 = first_frame，图生视频）或对象 {"url":"…","role":"first_frame|last_frame|reference_image"} |
video_urls |
array ≤3 | 参考视频 |
audio_urls |
array ≤3 | 参考音频；不能单独使用，必须搭配 role:"reference_image" 的图片或 video_urls；首帧/尾帧图片不能与音频同时使用 |
generate_audio |
boolean | 输出是否带音频，默认 false |
cURL · 含参考图与音频
curl https://tryaiapi.com/videos/v1/videos/generations \
-H "Authorization: Bearer $TRYAIAPI_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "doubao-seedance-2.0",
"prompt": "产品展示视频，柔和灯光，镜头缓慢推进",
"duration": 5,
"resolution": "720p",
"image_urls": [
{"url": "https://example.com/product.jpg", "role": "reference_image"}
],
"audio_urls": ["https://example.com/bgm.mp3"],
"generate_audio": true
}'
计费：按任务结算；成功后按上游 usage.total_tokens 精算，多退少补；失败全额退款不收费。
cURL · 工具调用（Function Calling）
curl https://tryaiapi.com/v1/chat/completions \
-H "Authorization: Bearer $TRYAIAPI_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "claude-sonnet-4-6",
"messages": [{"role": "user", "content": "杭州天气怎么样？"}],
"tools": [{
"type": "function",
"function": {
"name": "get_weather",
"parameters": {
"type": "object",
"properties": {"city": {"type": "string"}}
}
}
}]
}'
Python · 工具调用
from openai import OpenAI
client = OpenAI(base_url="https://tryaiapi.com/v1", api_key="$TRYAIAPI_KEY")
resp = client.chat.completions.create(
model="claude-sonnet-4-6",
messages=[{"role": "user", "content": "杭州天气怎么样？"}],
tools=[{
"type": "function",
"function": {
"name": "get_weather",
"parameters": {
"type": "object",
"properties": {"city": {"type": "string"}},
},
},
}],
)
print(resp.choices[0].message.tool_calls)
JavaScript · 工具调用
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://tryaiapi.com/v1",
apiKey: process.env.TRYAIAPI_KEY,
});
const resp = await client.chat.completions.create({
model: "claude-sonnet-4-6",
messages: [{ role: "user", content: "杭州天气怎么样？" }],
tools: [{
type: "function",
function: {
name: "get_weather",
parameters: {
type: "object",
properties: { city: { type: "string" } },
},
},
}],
});
console.log(resp.choices[0].message.tool_calls);
平台只提供官方标准模型 ID，不再提供 -thinking/-high/-low 等后缀变体。推理增强通过 API 参数 控制——Anthropic 用 thinking 对象，OpenAI 用 reasoning_effort。
Anthropic Claude · 启用 Extended Thinking
curl https://tryaiapi.com/v1/messages \
-H "Authorization: Bearer $TRYAIAPI_KEY" \
-H "Content-Type: application/json" \
-H "anthropic-version: 2023-06-01" \
-d '{
"model": "claude-opus-4-7",
"max_tokens": 16000,
"thinking": {
"type": "enabled",
"budget_tokens": 10000
},
"messages": [{
"role": "user",
"content": "解释量子纠缠的机制"
}]
}'
# thinking.budget_tokens 控制推理深度（1024 ~ 100000）
# max_tokens 须 > budget_tokens，留足 output 空间
# 支持 claude-opus-4-7, claude-opus-4-8 等支持 extended thinking 的模型
Python · Anthropic SDK
import anthropic
client = anthropic.Anthropic(
base_url="https://tryaiapi.com",
api_key="$TRYAIAPI_KEY",
)
resp = client.messages.create(
model="claude-opus-4-7",
max_tokens=16000,
thinking={
"type": "enabled",
"budget_tokens": 10000,
},
messages=[{"role": "user", "content": "解释量子纠缠的机制"}],
)
for block in resp.content:
if block.type == "thinking":
print("[思考过程]", block.thinking[:200], "...")
elif block.type == "text":
print("[回答]", block.text)
OpenAI · reasoning_effort 参数
curl https://tryaiapi.com/v1/chat/completions \
-H "Authorization: Bearer $TRYAIAPI_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5.5",
"reasoning_effort": "high",
"messages": [{
"role": "user",
"content": "解释量子纠缠的机制"
}]
}'
# reasoning_effort: "low" | "medium" | "high"（默认 medium）
# 适用于 gpt-5.5 等支持推理的模型；查 /v1/models 确认当前可用 ID
Python · OpenAI SDK
from openai import OpenAI
client = OpenAI(base_url="https://tryaiapi.com/v1", api_key="$TRYAIAPI_KEY")
resp = client.chat.completions.create(
model="gpt-5.5",
reasoning_effort="high",
messages=[{"role": "user", "content": "解释量子纠缠的机制"}],
)
print(resp.choices[0].message.content)
SDK Configuration
各客户端配置示例
Claude Code
Claude 全系# 设置环境变量后直接启动
export ANTHROPIC_BASE_URL=https://tryaiapi.com
export ANTHROPIC_API_KEY=your-tryaiapi-key
claude
Cursor
Claude + ChatGPT# Settings → Models → Override OpenAI Base URL
Base URL: https://tryaiapi.com
API Key: your-tryaiapi-key
# Claude 可直接填 claude-sonnet-4-6；ChatGPT 模型 ID 先查 /v1/models
Anthropic Python SDK
Claude 全系import anthropic
client = anthropic.Anthropic(
base_url="https://tryaiapi.com",
api_key="your-tryaiapi-key",
)
msg = client.messages.create(
model="claude-sonnet-4-6",
max_tokens=1024,
messages=[{"role": "user", "content": "Hello"}],
)
OpenAI Python SDK
Claude + ChatGPT# model 换名即切 provider
from openai import OpenAI
client = OpenAI(
base_url="https://tryaiapi.com/v1",
api_key="your-tryaiapi-key",
)
# Claude
resp = client.chat.completions.create(
model="claude-sonnet-4-6",
messages=[{"role": "user", "content": "Hello"}],
)
# ChatGPT — 同一个 client，具体模型 ID 先查 /v1/models
resp = client.chat.completions.create(
model="gpt-5.5",
messages=[{"role": "user", "content": "Hello"}],
)
视觉模型 API
Seedance / 视频任务# 标准 HTTP 调用：提交任务后轮询 poll_url
curl https://tryaiapi.com/videos/v1/videos/generations \
-H "Authorization: Bearer your-tryaiapi-key" \
-H "Content-Type: application/json" \
-d '{"model":"doubao-seedance-2.0","prompt":"A product video","duration":5,"resolution":"480p","aspect_ratio":"16:9"}'
Codex CLI
Claude + ChatGPT# Codex CLI 使用 OpenAI 兼容格式
export OPENAI_BASE_URL=https://tryaiapi.com/v1
export OPENAI_API_KEY=your-tryaiapi-key
codex
LangChain
Claude 全系from langchain_anthropic import ChatAnthropic
llm = ChatAnthropic(
base_url="https://tryaiapi.com",
api_key="your-tryaiapi-key",
model="claude-sonnet-4-6",
)
FAQ
常见问题
如何验证我用的是真模型？
每次请求的响应会包含上游官方的原始字段（Anthropic 的
request_id、OpenAI 的 id 和 system_fingerprint），与直连 API 完全一致。你也可以访问 透明度页面 查看数据库 schema，确认我们不存储任何 prompt 或 response 内容。
计费方式是什么？
文本模型按 token 用量计费；视觉模型按任务、时长和上游模型计费。登录控制台后可查看实时余额、用量明细、按模型/日期/Key 分组导出。支持对公转账 + 增值税发票。
你们会存储我的 prompt 和 response 吗？
文本通道不在普通数据库层保存 prompt 或 response 正文。视觉模型属于异步任务，任务参数、状态、成本、结果地址和失败原因会用于任务追踪与计费；完整边界以 透明度页面 为准。
支持 Streaming 吗？
文本模型支持 Anthropic 原生 SSE streaming 和 OpenAI 兼容 streaming，在请求中设置
"stream": true 即可。视觉模型不是 streaming 响应，而是提交任务后通过 /videos/v1/videos/generations/{task_id} 轮询结果。
视觉模型怎么调用？
使用同一个 API Key 调用
POST /videos/v1/videos/generations，提交后会返回 job_id 和 poll_url。任务完成后可在轮询响应或「视觉模型使用详情」里查看结果、下载地址和失败原因。
如何启用推理增强（Extended Thinking / reasoning_effort）？
平台只提供官方标准模型 ID，不提供
-thinking/-high/-low 等后缀变体。推理强度由 API 参数控制：
- Anthropic Claude：在请求体加
"thinking": {"type": "enabled", "budget_tokens": 10000}（预算 1024～100000 tokens）。max_tokens须大于budget_tokens。支持模型：claude-opus-4-7、claude-opus-4-8等。 - OpenAI（推理模型）：在请求体加
"reasoning_effort": "high"（可选low/medium/high）。支持模型：gpt-5.5等，查GET /v1/models确认。
和直连 Anthropic / OpenAI API 有什么区别？
文本通道保留 Anthropic / OpenAI 的兼容协议能力，包括 Streaming 和 Tool Calling；视觉模型由 Try AI API 的异步任务接口统一多家视频上游。区别是：同一个 Key、余额和使用记录打通，支持人民币对公结算 + 增值税发票。