← Managed inference

GLM-5.3

Complex coding and long-horizon agentic work

$1 in free creditsInput $1.40/1M · Output $4.40/1M

Overview

GLM-5.3 is Z.ai's open-weights reasoning model, aimed at complex coding and long-horizon tasks. It shares its base model with GLM-5.2, and Z.ai attributes every change to post-training. It reasons before answering, and this deployment returns the reasoning and any tool calls in their own response fields.

Strengths

  • Built for complex coding and long-horizon tasks
  • Tool calling alongside separately returned reasoning
  • 1M-token context window

Use cases

  • Coding agents with many tool calls per task
  • Long multi-step tasks that must stay on track

Good to know

Text input only on this alias. Served from a third-party 4-bit quantisation (W4AFP8: 4-bit weights, FP8 activations) of Z.ai's weights rather than the official checkpoint, so output can differ slightly from Z.ai's own API. Requests over 250K tokens run on a smaller long-context tier and can wait longer to start.

API access

One OpenAI-compatible API for every model in the catalog, plus the CLI and the function SDK inside workflows. Sign up and try $1 of inference free, then get $10 more once you've spent $5, and these requests work:

glm-5.3 · curl
curl https://model.graphn.ai/v1/chat/completions \
  -H "Authorization: Bearer $GRAPHN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "glm-5.3",
    "messages": [{"role": "user", "content": "Hello"}]
  }'