Add a choice 2/4

Showing 8 of 596. Type to narrow. Up to 4 choices side by side.

WHAT THE EVIDENCE SAYS

GLM-5.2 has the lowest estimated workload cost: $148.20 per month.

Editorial coverage is incomplete. Unscored models are not assigned invented scores.

Latency and reliability are not ranked without comparable measured conditions.

Workload assumptions 100 workloads / month

Text-token estimates only. Cache creation, cache storage, tools, taxes, and media are excluded. Without a documented cache-read rate, input is charged at the full listed rate.

Marks a row where choices differ. Unknown means unknown, not zero.
Decision criteriaGemini 4 HydrogenGoogle · API modelGLM-5.2Z.ai · API model
Overview
TypeAPI modelAPI model
ProviderGoogleZ.ai
Best suited forAccepts audio, files, images, text, video and produces text. Supports tool calling and reasoning controls.Teams that want an MIT-licensed, self-hostable model with good design taste and instant responses
Fit and cost
Saved editorial task fitNot scored75 / 100 · editorial
Monthly workload estimate$200.25$148.20
Price basisverified · checked Nov 14, 2026verified · checked Nov 14, 2026
Capability and access
Context window1,048,576 tokens1,048,576 tokens
Maximum output943,718 tokensn.a. on fetched pages
License / weightsNot documentedOpen weights
AccessOpenRouter API; availability and price can vary by routed provider.Z.ai API, GLM Coding Plan, Alibaba Cloud Model Studio (multi-region), Hugging Face weights
Evidence and limits
Important limitationsNo normalized editorial score or independently checked benchmark is assigned to this catalogue addition.GLM-5.3 costs the same and is much stronger (DeepSWE 46.2 -> 66.9) · DeepSWE 46.2 and Terminal Bench 3.0 4.6 show weak long-horizon agentic coding · Text-only on the evidence available · AA non-reasoning row collapses to 26
Source links2 linked sources8 linked sources
Benchmark observations0 observations; inspect conditions4 observations; inspect conditions
Inspect sourcesFull profile & evidence Full profile & evidence
Editorial assessmentVendor / source reportMeasured observationCost estimate

These evidence types are not interchangeable. Benchmark conditions and pricing sources are available in each profile.