Qwen3 Thinking 2507 API

alibaba/qwen3-235b-a22b-thinking-2507
Alibaba's Qwen 3 thinking 2507 is a multilingual reasoning model with knowledge augmentation, and creative capabilities.
Context
32K tokens
Input
$0.299 / 1M tokens
Output
$2.99 / 1M tokens
Released
Sep 9, 2025

How to use Qwen3 Thinking 2507 API

Install any OpenAI-compatible SDK, point it at api.aimlapi.com/v1, and set the model to alibaba/qwen3-235b-a22b-thinking-2507.
import requests

r = requests.post(
    "https://api.aimlapi.com/v1/chat/completions",
    headers={"Authorization": "Bearer " + AIMLAPI_KEY},
    json={
      "model": "alibaba/qwen3-235b-a22b-thinking-2507",
      "messages": [
        {
          "role": "user",
          "content": "Hello!"
        }
      ]
    },
)
print(r.json())
const r = await fetch("https://api.aimlapi.com/v1/chat/completions", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.AIMLAPI_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "model": "alibaba/qwen3-235b-a22b-thinking-2507",
    "messages": [
      {
        "role": "user",
        "content": "Hello!"
      }
    ]
  }),
});
console.log(await r.json());
curl -X POST https://api.aimlapi.com/v1/chat/completions \
  -H "Authorization: Bearer $AIMLAPI_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"alibaba/qwen3-235b-a22b-thinking-2507","messages":[{"role":"user","content":"Hello!"}]}'

OpenAI-compatible — swap the base URL and it works with your existing SDK.

Qwen3 Thinking 2507 API Pricing

TypePrice
Input
$0.299 / 1M tokens
Output
$2.99 / 1M tokens

Qwen3 Thinking 2507 Benchmarks

BenchmarkScoreWhat it measuresSourceRetrieved
LiveCodeBench
74.1%
Contamination-free competitive programming problemsSourceJuly 8, 2026
GPQA Diamond
81.1%
Google-proof graduate science questions (hardest subset)SourceJuly 8, 2026
AIME 2025
92.3%
Competition mathematics (AIME), 2025SourceJuly 8, 2026
Intelligence
12.7
Composite score across standardised reasoning, knowledge and problem-solving evaluations, measured independently by Artificial AnalysisSourceSeptember 12, 2026
Coding
22.1
Composite score across standardised coding evaluations, measured independently by Artificial AnalysisSourceSeptember 12, 2026
Math
91
Composite score across standardised mathematics evaluations, measured independently by Artificial AnalysisSourceSeptember 12, 2026

Qwen3 Thinking 2507 vs other models

ModelInputOutputContextBest for
$0.299 / 1M tokens
$2.99 / 1M tokens
32K tokens
Coding + agents
$6.5 / 1M tokens
$39 / 1M tokens
1.05M tokens
Reasoning + agents
$2.6 / 1M tokens
$13 / 1M tokens
1M tokens
Balanced coding + agents
$3.9 / 1M tokens
$19.5 / 1M tokens
1M tokens
Long-context, multimodal & agentic workflows
$0.65 / 1M tokens
$3.9 / 1M tokens
1.05M tokens
Reasoning + agents

Frequently asked questions

Qwen3 Thinking 2507 has a 32,000 tokens context window and can return up to 16,000 tokens.

Qwen3 Thinking 2507 takes text as input and returns text.

Use alibaba/qwen3-235b-a22b-thinking-2507 as the model id. Requests go to https://api.aimlapi.com/v1/chat/completions.

Qwen3 Thinking 2507 became available on September 9, 2025.

Qwen3 Thinking 2507 is priced at input $0.299 / 1M tokens, output $2.99 / 1M tokens.

Yes, Qwen3 Thinking 2507 can stream responses as they are generated.

Qwen3 Thinking 2507 was built by Alibaba Cloud.

Yes, it supports streaming responses along with tool use, including parallel tool calls.

Start building with Qwen3 Thinking 2507

Get API Key
1000+ models, one API.