Nexevo.aiNexevo.ai
← All examples
Agent / Automation

Multi-modal combo — research + math + image generation

One task chains web_search + python_exec + generate_image; the agent orchestrates steps automatically.

Python
python
from nexevo_ai import Nexevo
client = Nexevo()

# Multi-modal combo agent — 联网搜 → 算价格表 → 出封面图。
# 内部 4-6 步:
#   1. tool_call web_search(query="2026 latest LLM gateway competitors")
#   2. tool_call python_exec(expression="...")  # 算价格平均/中位
#   3. tool_call generate_image(prompt="...")  # 生封面海报
#   4. final_answer 文字总结 + 图片 URL

job = client.agents.run(
    task=(
        "调研 2026 H1 主流 LLM 网关 SaaS(OpenRouter / Together / "
        "Nexevo / Helicone),搜实时数据 + 算他们的平均 cost / 1M token,"
        "然后 generate_image 给我出一张极简对比海报。"
    ),
    tools=["web_search", "python_exec", "generate_image"],
    model="nexevo/balanced",
    max_steps=8,
)

print(job["result"])
# Step 1 → Step 2 → Step 3 → Step 4 自动串起来,
# 你只看 final_answer 就能拿到:实时调研 + 数字 + 海报 三合一交付。

# 流式版 — 实时看每步执行
for ev in client.agents.run_stream(task="...", tools=["web_search", "python_exec", "generate_image"]):
    if ev["type"] == "step":
        print(f"[{ev['step']['type']}]", ev['step'].get("tool_name", ""), ev['step'].get("content", "")[:80])
Multi-modal combo — research + math + image generation — Nexevo Cookbook | Nexevo.ai