MDL
GLM-5.2
Z AI
Verified against Artificial Analysis · Jul 27, 2026
Capability
66
percentile index
Price /M
$1.351 / $4.25
in / out
Context
1M
tokens
Avg score
37.6
4 benchmarks
AA Intelligence
34.1
index
Speed
86 tok/s
TTFT 1.9s
Released
Jun 16, 2026
About
GLM-5.2 is a open-weights model in the GLM family from Z AI. Benchmarked on 4 evals, averaging 37.6.
Benchmark scores · 4
avg 37.6 — higher is betterThe research behind this model
all papers mentioning it →LongStraw: Long-Context RL Beyond 2M Tokens under a Fixed GPU Budget▲ 201 on HF · Jul 16, 2026Single-Rollout Asynchronous Optimization for Agentic Reinforcement Learning▲ 26 on HF · Jul 8, 2026
Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.
Performance & price
Output speed86 tok/s
Latency (TTFT)1.9s
Price in / out$1.351 / $4.25 per 1M
Blended price$2.08 / 1M
Coding index46.5
Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →
History
full ledger →Capability over time
GPQA Diamondfirst seen at 68.6
15d ago
Humanity's Last Examfirst seen at 8.4
15d ago
SciCodefirst seen at 36.1
15d ago
AA Long-Context Reasoningfirst seen at 37.3
15d ago
Append-only ledger — every observed change to this model's numbers.
API providers
Artificial Analysis$4.25/M