Compare AI models
| Specifications | ||
ParametersThe learned values that shape how a model responds. B means billion and T means trillion. For mixture-of-experts models, the total count is larger than the number active for each token. Parameter count alone does not measure quality. | 16B | 744B |
Context windowHow much information the model can work with in one request, measured in tokens — pieces of words or other data. This includes your question, attachments, conversation and reply. Limits can vary by model variant and provider. | 8k | 1M |
| API pricingUSD per 1M tokens · lower wins | ||
Input priceWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it. | — | $1.40 |
Output priceWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one. | — | $4.40 |
Cached input priceA reduced rate for text you send over and over. If every request starts with the same instructions or the same document, the provider keeps a copy ready and charges less to read it again. | — | $0.26 |
Cheapest inputLowest input rate across every provider on OpenRouter, the lab included. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | — | $0.0066Relace |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | — | $1.44DeepInfra |
| — | $203DeepInfra · fp4 | |
| Benchmarks | ||
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 35.2% | 91.2% |
BullshitBench v2Nonsense detection — Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | — | 31% |
ProgramBenchProgram reconstruction — The AI receives a working program and its documentation, then builds a replacement from scratch without the original source code, internet access or decompilation. The score is the percentage of 200 programs that pass every behavioral test. We record each model's best published mini-SWE-agent result, including higher reasoning efforts where available. Partial test-pass rates and almost-solved programs do not count toward this score. Equal scores share a rank here; the official board also uses partial progress to break ties. Higher is better. | — | 0% |
SWE-Bench ProAgentic coding — Can the AI fix real bugs in real software? It's handed actual problems from open-source projects and has to write code that genuinely solves them. Higher is better. | — | 62.1% |
DeepSWE 1.1Agentic coding — Artificial Analysis' independent test of deep, agentic software-engineering work — the AI has to plan and carry out substantial coding tasks end to end. (Version 1.1 of the test.) Higher is better. | — | 44% |
Next.js EvalsNext.js coding — Vercel's open eval of how well AI coding agents build and migrate real Next.js apps — measured as the share of tasks the agent completes successfully. Higher is better. | — | 82% |
NL2Repo-BenchRepo-level code generation — Tests whether the AI can turn a natural-language requirement into working code across an entire repository, not just produce a single function or patch. Higher is better. | — | 48.9% |
LiveCodeBenchCompetitive coding — Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better. | 14.3% | — |
HumanEvalFunction synthesis — 164 small Python problems: the AI is given a function's description and has to write the working function. This was the coding benchmark of the GPT-3.5 and GPT-4 era, before the field moved to fixing real bugs in real repositories. Higher is better. | 48.1% | — |
Frontier-Bench v0.1Agentic computer work — A hard, ever-evolving set of real computer tasks — coding, system administration, data work, and more — that an AI agent has to complete on its own. Run by the Harbor / Laude Institute team as the successor to Terminal-Bench (v0.1 is the first release of the task set). The score is the share of tasks solved. Higher is better. | — | 5.1% |
Terminal-Bench 2.1Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better. | — | 81% |
Humanity's Last Exam · text only, no toolsMultidisciplinary reasoning — accuracy on the 2,158 text-only questions without tools | — | 40.5% |
Humanity's Last Exam · with toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects. “With tools” means the AI is allowed to search the web or run code while answering. Higher is better. | — | 54.7% |
MATHReasoning — accuracy on the original competition mathematics test set | 45.3% | — |
WinoGrandeReasoning — resolving ambiguous pronouns in context | 74.6% | — |
AIME 2026Reasoning — single-answer accuracy on the 2026 invitational mathematics exam without tools | — | 99.2% |
SimpleQAKnowledge — accuracy on short, fact-seeking questions in the original SimpleQA set | 5.6% | — |
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | 42.4% | — |
MMLUGeneral knowledge — A 57-subject multiple-choice exam — history, law, medicine, maths — that was the standard measure of how much a model knows from 2020 until roughly 2024, when frontier scores crowded into the high 80s and labs moved on to harder tests. The scores here were published years apart under different testing setups, so read them as a historical record rather than a like-for-like ranking. Higher is better. | 70% | — |
GSM8KGrade-school math — Grade-school maths word problems that take a few steps of arithmetic to work through. It separated the models of 2022 and 2023 sharply, then saturated. One caveat on the historical numbers: OpenAI included part of the GSM8K training set in GPT-4's pre-training mix, so GPT-4's score is not a clean few-shot result. Higher is better. | 77.4% | — |
GDPval-AA v2Knowledge work — economically valuable knowledge work (v2, re-based Elo) | — | 1514 |
| Overview | ||
| Company | Moonshot AI | Z.ai |
| Release date | Feb 22 2025 | Jun 16 2026 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
MoonlightvsClaude Fable 5.2GLM-5.2vsClaude Fable 5.2MoonlightvsGPT-6.2 AstraGLM-5.2vsGPT-6.2 AstraMoonlightvsGemini 4 HydrogenGLM-5.2vsGemini 4 HydrogenMoonlightvsMuse EmberGLM-5.2vsMuse EmberMoonlightvsGrok 4.7GLM-5.2vsGrok 4.7MoonlightvsDeepSeek-V4.1-Flash-1017GLM-5.2vsDeepSeek-V4.1-Flash-1017Frequently asked questions
GLM-5.2 leads Moonlight on 1 of the 1 benchmark they both report (GPQA Diamond). Only GLM-5.2 has a verified first-party price for an available API: $1.40 per million input tokens and $4.40 per million output tokens. No currently available pay-as-you-go API rate is tracked for Moonlight. Moonlight shipped 479 days before GLM-5.2, so benchmark comparisons should account for the intervening progress.
Moonlight has 16B parameters, while GLM-5.2 has 744B. Context windows are 8k (Moonlight) vs 1M (GLM-5.2).
On GPQA Diamond, GLM-5.2 leads at 91.2% vs Moonlight at 35.2%.