GLM-5 API: pricing, base URL and quick setup

Specifications and pricing

Prices synced with platform billing (USD per 1M tokens).

Quick setup

curl https://a2agent.me/v1/chat/completions \
  -H "content-type: application/json" \
  -H "authorization: Bearer sk-your-a2agent-key" \
  -d '{
    "model": "glm-5",
    "messages": [{"role": "user", "content": "Reply with OK"}]
  }'

Using it from a coding client

Agent frameworks that fan out many small tool calls fit GLM-5 well, and Cline keeps each request far enough under 200K that the ceiling rarely binds. If a session keeps hitting it, move to GLM-5.3 rather than trimming prompts.

• Run it in Cline
• Run it in Claude Code

How it compares within its family

GLM-5 is the agent-tagged member of the Z.ai family and the cheapest of the 5.x releases sold here, at $1.00 per million input tokens and $3.20 output. It is also the shortest: 200K, against 1M on GLM-5.2, GLM-5.3 and GLM-5.3 Flash. GLM-5.1 shares the same 200K window but is tagged for coding and lists a little higher, at $1.15 and $4.00. Inside Z.ai the read is therefore straightforward. GLM-5 handles orchestration and tool loops. GLM-5.1 handles code at the same window size. The 5.2 and 5.3 releases are what you move to when a prompt outgrows 200K. Going up the line buys context, not a cheaper token.

Other models on this platform

• GLM-5.1200K
• GLM-5.21M
• GLM-5.31M
• GLM-5.3 Flash1M

Questions