Two new models from the Qwen3.7 family are live on Antbase, routed through NetMind. They're built for different jobs: Qwen3.7-Max is the flagship for reasoning and long-running agents, and Qwen3.7-Plus is a lower-cost multimodal model that also reads images and video. Both run on a 1M-token context, so you can hold a whole codebase or a long document in view — and you switch between them by changing a single model name.
Two tiers, one model string away
- •Qwen3.7-Max — reasoning and long-horizon agents: frontier coding, sustained autonomous execution, MCP and multi-agent orchestration. Text in, reasoning out.
- •Qwen3.7-Plus — multimodal at lower cost: full-modality input (text + image + video), visual agent perception/grounding, GUI and CLI operation.
Because Antbase is OpenAI-compatible, picking a tier is a one-token change — your SDK, prompts, tool definitions, and agent loop stay exactly the same:
from openai import OpenAI
client = OpenAI(base_url="https://antbase.ai/v1", api_key="ant-...")
# Reasoning + long-horizon agents
resp = client.chat.completions.create(
model="netmind/Qwen/Qwen3.7-Max",
messages=[{"role": "user", "content": "Plan the refactor, then execute it."}],
)
# Same call, multimodal + cheaper — just change the model name
resp = client.chat.completions.create(
model="netmind/Qwen/Qwen3.7-Plus",
messages=[{"role": "user", "content": "Describe what's happening in this screen recording."}],
)Pricing (via NetMind), per 1M tokens
- •Qwen3.7-Max — input $1.644, output $4.932, cache read $0.329. One rate across the full 1M context.
- •Qwen3.7-Plus (up to 256K) — input $0.274, output $1.096, cache read $0.0548.
- •Qwen3.7-Plus (256K–1M) — input $0.822, output $3.288, cache read $0.164.
These route through Antbase to NetMind, so you get the pricing above behind your existing Antbase key — same dashboard, same usage tracking, same fallback behavior as every other model you call.
Find both in the model catalog:
Or drop them straight into the playground and compare the two tiers on the same prompt:
Put Qwen3.7-Max and Qwen3.7-Plus into your agent, switch tiers with a single model string, and let Antbase handle the routing.
