Qwen3.8-27BvsGLM-5.3-Flash
Qwen3.8-27B | GLM-5.3-Flash | |
|---|---|---|
| Specifications | ||
ParametersA rough measure of how big the model is. More parameters usually means more capable and more expensive to run, though it is a poor guide on its own — a smaller, newer model often beats a larger, older one. | 27B | 320B |
Context windowHow much text the model can hold in mind at once — your question, any documents you attach, the conversation so far, and its own reply. Go past it and the earliest part falls out of view. | 262k | 1M |
| API pricingUSD per 1M tokens · lower wins | ||
Cheapest inputLowest input rate across third-party providers, excluding the lab itself. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | $0.35Chutes | — |
Cheapest outputLowest output rate across third-party providers, excluding the lab itself. May come from a different provider than the cheapest input. | $2.55AkashML | — |
| Benchmarks | ||
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. | 42.2% | 63.4% |
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. | 73% | 84.3% |
Agent's Last Exam · pass@1Agentic computer use — A hard set of desktop and operating-system tasks an AI agent has to finish by looking at the screen and working the machine itself. The score is the share it passes outright — partial credit does not count. Higher is better. | 20.4% | 26.3% |
| BenchmarksPublished by one model only | ||
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. | 78% | — |
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. | 61.7% | — |
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. | 42.3% | — |
QwenSWEBenchSoftware engineering — Qwen's in-house coding benchmark for evaluating a model's ability to complete software-engineering work. Higher is better. | 79% | — |
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. | 90.3% | — |
JobBenchProfessional tool use — Tests the AI on professional workplace tasks that require using real work tools — the kind of multi-step jobs an office worker handles. Higher is better. | 33.4% | — |
CoWorkBenchLong-horizon office work — Tests long-running office tasks across fields including computer science, finance, law, medicine, and other productivity work. Higher is better. | 70.7% | — |
Humanity's Last Exam · no toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects, written so you can't just look up the answer. “No tools” means the AI answers on its own. Higher is better. | 30.8% | — |
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. | — | 55.3% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 89.2% | — |
IFBenchInstruction following — Tests whether the AI can follow detailed instructions and satisfy multiple constraints at once. Higher is better. | 79.5% | — |
OSWorld-VerifiedAgentic computer use — Can the AI actually operate a computer — clicking, typing, and using real apps — to finish tasks on its own? Higher is better. | 84.3% | — |
Agent's Last Exam · scoreAgentic computer use — The graded score on the same desktop and operating-system tasks in Agent's Last Exam, giving partial credit for progress beyond the strict pass-or-fail result. Higher is better. | 42.9% | — |
AutomationBenchBusiness workflows — Tests whether the AI can run real multi-step business workflows — the kind of end-to-end office processes companies want to automate — from start to finish. Higher is better. | — | 48.8% |
GDPval-AA v2Knowledge work — economically valuable knowledge work (v2, re-based Elo) | — | 1773 |
| Overview | ||
| Company | Qwen | Z.ai |
| Release date | Aug 14 2026 | Aug 26 2026 |
| Access | Open Weight | Open Weight |
Other comparisons
Qwen3.8-27BvsClaude Opus 5GLM-5.3-FlashvsClaude Opus 5Qwen3.8-27BvsGPT-5.6 SolGLM-5.3-FlashvsGPT-5.6 SolQwen3.8-27BvsGemini 3.7 FlashGLM-5.3-FlashvsGemini 3.7 FlashQwen3.8-27BvsMuse GlimmerGLM-5.3-FlashvsMuse GlimmerQwen3.8-27BvsGrok 4.6GLM-5.3-FlashvsGrok 4.6Qwen3.8-27BvsDeepSeek-V4-Pro-0813GLM-5.3-FlashvsDeepSeek-V4-Pro-0813
Frequently asked questions
GLM-5.3-Flash leads Qwen3.8-27B on 3 of the 3 benchmarks they both report (DeepSWE 1.1, Terminal-Bench 2.1, Agent's Last Exam). Qwen3.8-27B shipped 12 days before GLM-5.3-Flash, so benchmark comparisons should account for the intervening progress.
Qwen3.8-27B has 27B parameters, while GLM-5.3-Flash has 320B. Context windows are 262k (Qwen3.8-27B) vs 1M (GLM-5.3-Flash).
On DeepSWE 1.1, GLM-5.3-Flash leads at 63.4% vs Qwen3.8-27B at 42.2%. On Terminal-Bench 2.1, GLM-5.3-Flash leads at 84.3% vs Qwen3.8-27B at 73%. On Agent's Last Exam · pass@1, GLM-5.3-Flash leads at 26.3% vs Qwen3.8-27B at 20.4%.
Qwen3.8-27B was released by Qwen on Aug 14 2026.
GLM-5.3-Flash was released by Z.ai on Aug 26 2026.
GLM-5.3-Flash leads on DeepSWE 1.1 — Qwen3.8-27B 42.2% vs GLM-5.3-Flash 63.4%.
Qwen3.8-27B has a 262k context window; GLM-5.3-Flash has 1M.