Qwen 3.7 Max & Plus Are Now Available on TheRouter.ai

Update 2026-10-17: Alibaba retired qwen3-max, qwen3.6-max-preview, qwen3-coder-plus and qwen3-vl-flash on 2026-10-10, and TheRouter no longer serves them. Code examples below now use the live successors (qwen/qwen3.7-max, qwen/qwen3.7-plus, qwen/qwen3.7-flash).

Alibaba Cloud announced Qwen 3.7 Max and Qwen 3.7 Plus at the Alibaba Cloud Summit in Hangzhou on May 19–21, 2026. Both models are now live on TheRouter.ai under a single API key and endpoint.

The short version: Max is a reasoning-first text model that outperforms GPT-5.5 on SWE-Bench Pro. Plus is a multimodal variant — text, image, and video input — at about one-sixth the price. If you were already using qwen/qwen3-max (retired), read the comparison section below before switching.

Model Specs at a Glance

Max pricing reflects a promotional 50% discount off list price ($2.50/$7.50 per 1M tokens). Plus pricing is standard list.

Qwen 3.7 Max: Agent-First Reasoning

Qwen 3.7 Max is designed for agentic workloads: multi-step reasoning, code generation, and autonomous tool use. The benchmark headline is SWE-Bench Verified: 80.4%, which puts it ahead of GPT-5.5 on SWE-Bench Pro (60.6% vs 58.6%).

That's a meaningful delta on a benchmark that directly measures software engineering capability — fixing real GitHub issues in real codebases. If your primary use case is code agents, coding assistants, or automated PR workflows, the Max numbers are worth taking seriously.

Key specs:

• Context: 1,000,000 tokens
• Modality: text-only
• SWE-Bench Verified: 80.4%
• SWE-Bench Pro vs GPT-5.5: 60.6% vs 58.6%
• Architecture: reasoning-focused, agent-first design

At $1.25/$3.75 per 1M tokens (promotional), Max is priced competitively against comparable frontier reasoning models.

Qwen 3.7 Plus: Multimodal at Low Cost

Qwen 3.7 Plus covers text, image, and video input within the same 1M token context window. Visual tokens count against the shared context budget, so factor that in for heavy vision workloads with long conversation histories.

At $0.32/$1.28 per 1M tokens, Plus is roughly one-sixth the price of Max. For document understanding, image analysis, or multimodal RAG pipelines where you don't need Max-level reasoning depth, Plus is the cost-efficient choice.

Key specs:

• Context: 1M tokens (text + visual tokens combined)
• Modality: text, image URL, video input
• Input price: $0.32 / 1M tokens
• Output price: $1.28 / 1M tokens

Pricing Comparison

For a typical agentic batch job: 10M input tokens + 2M output tokens:

• Qwen 3.7 Max: $12.50 + $7.50 = $20.00
• Qwen 3.7 Plus: $3.20 + $2.56 = $5.76

Quick Start

Both models are available through the standard OpenAI-compatible endpoint. No special headers or API changes required — just update the model field.

curl:

curl https://api.therouter.ai/v1/chat/completions \
  -H "Authorization: Bearer $THEROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen/qwen3.7-max",
    "messages": [
      {
        "role": "user",
        "content": "Review this Python function for bugs and suggest improvements."
      }
    ],
    "max_tokens": 2048
  }'

Python (openai SDK):

from openai import OpenAI

client = OpenAI(
    api_key="your-therouter-api-key",
    base_url="https://api.therouter.ai/v1",
)

# Qwen 3.7 Max — text reasoning
response = client.chat.completions.create(
    model="qwen/qwen3.7-max",
    messages=[
        {"role": "user", "content": "Write a Python script to parse and validate JSON logs from a Kubernetes pod."}
    ],
    max_tokens=2048,
)
print(response.choices[0].message.content)

# Qwen 3.7 Plus — vision input
response = client.chat.completions.create(
    model="qwen/qwen3.7-plus",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "What's shown in this screenshot? Identify any UI issues."},
                {"type": "image_url", "image_url": {"url": "https://example.com/screenshot.png"}},
            ],
        }
    ],
    max_tokens=1024,
)
print(response.choices[0].message.content)

Max vs Plus vs qwen3-max: Which One to Use

If you're currently on qwen/qwen3-max (retired), here's the decision matrix:

Use qwen/qwen3.7-max if:

• Your workload is code-heavy or agent-driven
• You need the best reasoning performance available from Qwen
• The 80.4% SWE-Bench Verified score is relevant to your use case
• Budget is secondary to output quality

Use qwen/qwen3.7-plus if:

• You need image or video input
• Cost efficiency matters — it's ~6x cheaper than Max
• You're building multimodal RAG, document analysis, or visual understanding features
• Reasoning depth of Max isn't required

Staying on qwen3-max is no longer possible: Alibaba retired it on 2026-10-10. Move to qwen/qwen3.7-max.

In most new projects starting today, 3.7 Max or 3.7 Plus is the better default over the previous qwen3-max generation.

Access via TheRouter.ai

Both models are live now. Use your existing TheRouter API key at api.therouter.ai — no provisioning, no separate DashScope account, no region-specific endpoints to manage. TheRouter routes through Aliyun Bailian / DashScope and normalizes the response format to OpenAI-compatible output.

Model IDs:

• qwen/qwen3.7-max
• qwen/qwen3.7-plus