Sulat.com
AI models
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3 from Synthetic
CoralBricks logo

Model details

GLM 5.3 FP4

GLM 5.3 FP4 is positioned as a hosted option for agentic coding and other long-context text tasks. CoralBricks’ documentation demonstrates use through the Pi terminal coding harness, while its OpenAI-compatible endpoint makes the model straightforward to connect to compatible agent frameworks and existing API-based workflows. The documentation groups it with other open models, though the supplied evidence does not establish a specific open-weights license for this variant.

The model’s clearest practical advantage is its large working context, with CoralBricks listing the endpoint at up to 1M tokens. That capacity is well suited to code repositories, extended documents, and multi-step agent sessions that need a broad span of material available at once. No supplied evidence documents the architecture, training process, parameter count, quantization method, or benchmark performance, so those details should not be inferred from the model name or treated as confirmed.

CoralBricksglm-5.3-fp4glm

Quick Info

Powered by
Provider
CoralBricks
Model key
glm-5.3-fp4
Release date
Aug 14, 2026
Last updated
Aug 14, 2026
Input modalities
Output modalities
Capabilities

Cost

Input token cost
$1.12
Output token cost
$4.40

Limits

Output tokens
131,072 tokens
Context window
1,048,576 tokens

Latest news about GLM 5.3 FP4

We haven't found any news about GLM 5.3 FP4 yet. New models can take a few days to get coverage, so check back soon or look again now.

Videos about GLM 5.3 FP4

More models around GLM 5.3 FP4