Skip to content
Models · Jun 19, 2026

Z.ai releases GLM-5.2, a 753B-parameter open-weights text-only LLM with 1M token context

The model leads independent open-weights benchmarks but consumes more output tokens per task than peers.

Trust79
HypeLow hype

1 source · cross-referenced

ShareXLinkedInEmail
TL;DR
  • Chinese lab Z.ai released GLM-5.2 as open weights under an MIT license on June 16, 2026, following a subscriber preview on June 13.
  • The 753B-parameter MoE model has a 1M token context window and is ranked the leading open-weights model on the Artificial Analysis Intelligence Index v4.1.
  • GLM-5.2 uses more output tokens per task than other leading open-weights models, according to Artificial Analysis.
  • It ranks second on the Code Arena WebDev leaderboard for front-end web development tasks.

Chinese AI lab Z.ai released GLM-5.2 as open weights under an MIT license on June 16, 2026, after a subscriber preview on June 13. The model is a 753B-parameter mixture-of-experts text-only LLM with 40 active experts and a 1.51TB weight size.

GLM-5.2 expands the context window to 1 million tokens, up from 200,000 in GLM-5.1. It is positioned as a text-only model, while Z.ai maintains a separate vision family (e.g., GLM-5V-Turbo) that is not open weights.

Artificial Analysis reports GLM-5.2 as the leading open-weights model on the Intelligence Index v4.1 with a score of 51, ahead of MiniMax-M3 (44), DeepSeek V4 Pro (max, 44), and Kimi K2.6 (43).

The same evaluation found GLM-5.2 to be token-hungry, using 43k output tokens per task compared to 26k for GLM-5.1 and 24k–37k for peer models.

On the Code Arena WebDev leaderboard, which measures front-end web development tasks including agentic coding workflows, GLM-5.2 ranks second behind Claude Fable 5.

OpenRouter lists GLM-5.2 from nine providers, with most charging $1.40 per million input tokens and $4.40 per million output tokens.

Sources
  1. 01Simon Willison — everythingGLM-5.2 is probably the most powerful text-only open weights LLM
Also on Models

Stories may contain errors. Dispatch is assembled with AI assistance and curated by human editors; despite the trust-score filter, mistakes happen. We correct publicly — every article links to its revision history. Nothing here is financial, legal, or medical advice. Verify before relying on any claim.

© 2026 Dispatch. No ads. No sponsorships. No paid placement. Reader-supported via Ko-fi.

Built by a person who cares about honest AI news.