OPEN LINK ↗
// 1h agoINFRASTRUCTURE
Merge Gateway slashes GLM-5.3-Flash prices
Merge Gateway is offering Z.ai’s newly released GLM-5.3-Flash at $0.012/M input, $0.04/M output, and $0.003/M cached input through September. The 320B MoE model activates 18B parameters, supports native vision, tool calling, and a 1M-token context.
// ANALYSIS
This pricing makes GLM-5.3-Flash compelling for high-volume coding and agent workloads, though developers should treat the discount as temporary infrastructure, not a permanent cost baseline.
- –Promotional rates are roughly 90% below Z.ai’s standard API pricing.
- –Its Intelligence Index score of 57 puts it in serious competition with far more expensive models.
- –Sparse-plus-linear attention and 18B active parameters help explain the model’s serving-cost advantage.
- –Merge Gateway adds unified API access, routing, spend controls, and observability for production deployments.
- –Teams should test reliability, latency, rate limits, and fallback costs before building around the September promotion.
// TAGS
glm-5.3-flashllmopen-weightsmultimodalmoeinferenceapipricing
DISCOVERED
1h ago
2026-08-27
PUBLISHED
2h ago
2026-08-27
RELEVANCE
9/ 10
AUTHOR
merge_api