Compare / head-to-head
Gemini 3.8 FlashvsGLM-5.3-Flash
Gemini 3.8 Flash leads 5 of 5 shared benchmarks. GLM-5.3-Flash is 6.9x cheaper per token. Gemini 3.8 Flash has the larger context window (1.0M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
5 – 0
Gemini 3.8 Flash leads
Cheaper per token
GLM-5.3-Flash
6.9x cheaper, input + output
Larger context
Gemini 3.8 Flash
1.0M tokens
Provider uptime (30d)
100% · —
Google · Z.ai
Gemini 3.8 Flash
Google · released 2026-09-02
textvisionaudio
GLM-5.3-Flash
Z.ai · released 2026-09-16
textvision
Quality
Benchmark matrix
5 shared · 10 only Gemini 3.8 Flash · 14 only GLM-5.3-Flash| Benchmark | Gemini 3.8 Flash | GLM-5.3-Flash | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 5 | ||||
| GPQA Diamond | 95.3% ↗ | 90.2% ↗epoch | +5.1 pt | Gemini 3.8 Flash |
| DeepSWE v1.1 | 73.7% ↗ | 63.4% ↗ | +10.3 pt | Gemini 3.8 Flash |
| FrontierMath Tier 4 v2 (Epoch AI run) | 22% ↗epoch | 17.1% ↗epoch | +4.9 pt | Gemini 3.8 Flash |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 68.4% ↗epoch | 55.8% ↗epoch | +12.6 pt | Gemini 3.8 Flash |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 98.9% ↗epoch | 93.9% ↗epoch | +5 pt | Gemini 3.8 Flash |
| Only Gemini 3.8 Flash reports · 10 | ||||
| CharXiv Reasoning (no tools) | 86.2% ↗ | not reported | — | — |
| GDPVal-AA v2 | 1545 ↗ | not reported | — | — |
| Harvey's Legal Agent Benchmark | 10% ↗ | not reported | — | — |
| HLE-Verified | 54.9% ↗ | not reported | — | — |
| LABBench2 | 86.2% ↗ | not reported | — | — |
| OSWorld-2.0 | 59% ↗ | not reported | — | — |
| SimpleQA Verified | 69.7% ↗epoch | not reported | — | — |
| Terminal-bench 2.1 | 89.4% ↗ | not reported | — | — |
| Terminal-bench 4.0 | 19.1% ↗ | not reported | — | — |
| Vals Finance Agent v2 | 61.4% ↗ | not reported | — | — |
| Only GLM-5.3-Flash reports · 14 | ||||
| LMArena Elo | not reported | 1471.9 ↗ | — | — |
| 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% ↗ | — | — |
| 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% ↗ | — | — |
| 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 Gemini 3.8 Flash 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 | Gemini 3.8 Flash | GLM-5.3-Flash | Edge |
|---|---|---|---|
| Context window | 1.0M tokens | 1M tokens | Gemini 3.8 Flash |
| Max output | 66K tokens | 131K tokens | GLM-5.3-Flash |
| Input price / 1M | $0.75 ↗ | $0.15 ↗ | GLM-5.3-Flash |
| Output price / 1M | $3.75 ↗ | $0.5 ↗ | GLM-5.3-Flash |
| Cached input / 1M | $0.075 ↗ | $0.03 ↗ | GLM-5.3-Flash |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $4.5 | $0.65 | GLM-5.3-Flash |
| Modalities | text · vision · audio | text · vision | Gemini 3.8 Flash |
| Released | 2026-09-02 | 2026-09-16 | — |
| Cited benchmark scores | 16 | 19 | — |
Reliability
Provider status
Last 60 days · refreshed daily on this page · live on /statusMore matchups
Gemini 3.8 Flash 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 Gemini 3.8 Flash and GLM-5.3-Flash on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.