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. | — | 355B |
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. | 1M | 200k |
| 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. | — | $0.60 |
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. | — | $2.20 |
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. | — | $0.11 |
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. | — | $0.32DeepInfra |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | — | $1.40DeepInfra |
| — | $168DeepInfra · fp4 | |
| Benchmarks | ||
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. | 37.8% | 84.9% |
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | 64.5% | 84.3% |
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | — | 73.8% |
SWE-Bench MultilingualMultilingual coding — Like SWE-Bench, but the coding problems span many programming languages, not just one. Tests how broadly the AI can code. Higher is better. | — | 66.7% |
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. | 86.6% | — |
Terminal-Bench 2.0Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? (Version 2.0 of the test.) Higher is better. | — | 41% |
IFEvalAgentic tasks — following verifiable instructions in generated responses | 76.3% | — |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | — | 52% |
Humanity's Last Exam · text only, no toolsMultidisciplinary reasoning — accuracy on the 2,158 text-only questions without tools | — | 24.8% |
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. | — | 42.8% |
MATHReasoning — accuracy on the original competition mathematics test set | 81.1% | — |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | — | 95.7% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | — | 85.7% |
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. | 93.8% | — |
| Overview | ||
| Company | Qwen | Z.ai |
| Release date | Nov 15 2024 | Dec 22 2025 |
| Access | Closed | Open Weight |
| Model details | View model | View model |
Other comparisons
Qwen2.5-Turbo (1M)vsClaude Fable 5.2GLM-4.7vsClaude Fable 5.2Qwen2.5-Turbo (1M)vsGPT-6.2 AstraGLM-4.7vsGPT-6.2 AstraQwen2.5-Turbo (1M)vsGemini 4 NeonGLM-4.7vsGemini 4 NeonQwen2.5-Turbo (1M)vsMuse EmberGLM-4.7vsMuse EmberQwen2.5-Turbo (1M)vsGrok 4.7GLM-4.7vsGrok 4.7Qwen2.5-Turbo (1M)vsDeepSeek-V4.1-Flash-1017GLM-4.7vsDeepSeek-V4.1-Flash-1017Frequently asked questions
GLM-4.7 leads Qwen2.5-Turbo (1M) on 2 of the 2 benchmarks they both report (LiveCodeBench, MMLU-Pro). Only GLM-4.7 has a verified first-party price for an available API: $0.60 per million input tokens and $2.20 per million output tokens. No currently available pay-as-you-go API rate is tracked for Qwen2.5-Turbo (1M). Qwen2.5-Turbo (1M) shipped 402 days before GLM-4.7, so benchmark comparisons should account for the intervening progress.
Context windows are 1M (Qwen2.5-Turbo (1M)) vs 200k (GLM-4.7). Qwen2.5-Turbo (1M) is closed, while GLM-4.7 is open weight.
On LiveCodeBench, GLM-4.7 leads at 84.9% vs Qwen2.5-Turbo (1M) at 37.8%. On MMLU-Pro, GLM-4.7 leads at 84.3% vs Qwen2.5-Turbo (1M) at 64.5%.