Compare / head-to-head
GLM-4.6vsGLM-5.3-Flash
GLM-5.3-Flash leads 1 of 1 shared benchmark. GLM-5.3-Flash is 4.3x cheaper per token. GLM-5.3-Flash has the larger context window (1M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
0 – 1
GLM-5.3-Flash leads
Cheaper per token
GLM-5.3-Flash
4.3x cheaper, input + output
Larger context
GLM-5.3-Flash
1M tokens
Provider uptime (30d)
—
Z.ai
GLM-5.3-Flash
Z.ai · released 2026-09-16
textvision
Quality
Benchmark matrix
1 shared · 1 only GLM-4.6 · 18 only GLM-5.3-Flash| Benchmark | GLM-4.6 | GLM-5.3-Flash | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 1 | ||||
| LMArena Elo | 1440.5 ↗ | 1471.9 ↗ | -31.4 | GLM-5.3-Flash |
| Only GLM-4.6 reports · 1 | ||||
| AIME 2025 | 91.7% ↗matharena ⚠ | not reported | — | — |
| Only GLM-5.3-Flash reports · 18 | ||||
| GPQA Diamond | not reported | 90.2% ↗epoch | — | — |
| Agents' Last Exam (ALE-CLI) | not reported | 26.3% ↗ | — | — |
| AutomationBench v1.0.6 | not reported | 48.8% ↗ | — | — |
| BabyVision | not reported | 53.4% ↗ | — | — |
| Chartography (with tools) | not reported | 78% ↗ | — | — |
| CharXiv Reasoning (with tools) | not reported | 89.4% ↗ | — | — |
| DeepSWE v1.1 | not reported | 63.4% ↗ | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | not reported | 17.1% ↗epoch | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | not reported | 55.8% ↗epoch | — | — |
| GDPval-AA v2 | not reported | 1773 ↗ | — | — |
| HLE (with tools) | not reported | 55.3% ↗ | — | — |
| MMVU | not reported | 80.5% ↗ | — | — |
| MVBench | not reported | 77.8% ↗ | — | — |
| NL2Repo | not reported | 56.3% ↗ | — | — |
| OfficeQA Pro | not reported | 62.4% ↗ | — | — |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | not reported | 93.9% ↗epoch | — | — |
| Terminal-Bench 2.1 | not reported | 84.3% ↗ | — | — |
| Toolathlon Verified | not reported | 78.4% ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is GLM-4.6 minus GLM-5.3-Flash in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | GLM-4.6 | GLM-5.3-Flash | Edge |
|---|---|---|---|
| Context window | 200K tokens | 1M tokens | GLM-5.3-Flash |
| Max output | 131K tokens | 131K tokens | Tie |
| Input price / 1M | $0.6 ↗ | $0.15 ↗ | GLM-5.3-Flash |
| Output price / 1M | $2.2 ↗ | $0.5 ↗ | GLM-5.3-Flash |
| Cached input / 1M | $0.11 ↗ | $0.03 ↗ | GLM-5.3-Flash |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $2.8 | $0.65 | GLM-5.3-Flash |
| Modalities | text | text · vision | GLM-5.3-Flash |
| Released | — | 2026-09-16 | — |
| Cited benchmark scores | 2 | 19 | — |
Reliability
Z.ai status
Last 60 days · refreshed daily on this page · live on /statusMore matchups
GLM-4.6 vs …
Models sharing the most benchmarksMore matchups
GLM-5.3-Flash vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run GLM-4.6 and GLM-5.3-Flash on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.