Compare / head-to-head
Gemini 3.1 Flash-LitevsGLM-5.3
GLM-5.3 leads 3 of 4 shared benchmarks. Gemini 3.1 Flash-Lite is 3.3x cheaper per token. Gemini 3.1 Flash-Lite has the larger context window (1.0M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
1 – 3
GLM-5.3 leads
Cheaper per token
Gemini 3.1 Flash-Lite
3.3x cheaper, input + output
Larger context
Gemini 3.1 Flash-Lite
1.0M tokens
Provider uptime (30d)
100% · —
Google · Z.ai
Gemini 3.1 Flash-Lite
Google · released 2026-03-03
textvisionaudio
Quality
Benchmark matrix
4 shared · 5 only Gemini 3.1 Flash-Lite · 20 only GLM-5.3| Benchmark | Gemini 3.1 Flash-Lite | GLM-5.3 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 4 | ||||
| GPQA Diamond | 86.9% ↗ | 90.9% ↗epoch | -4 pt | GLM-5.3 |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 27.7% ↗epoch | 68.8% ↗epoch | -41.1 pt | GLM-5.3 |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 80% ↗epoch | 91.1% ↗epoch | -11.1 pt | GLM-5.3 |
| SimpleQA Verified | 43.3% ↗ | 41% ↗epoch | +2.3 pt | Gemini 3.1 Flash-Lite |
| Only Gemini 3.1 Flash-Lite reports · 5 | ||||
| Humanity's Last Exam (no tools) | 16% ↗ | not reported | — | — |
| LiveCodeBench | 72% ↗ | not reported | — | — |
| MMMLU | 88.9% ↗ | not reported | — | — |
| MMMU-Pro | 76.8% ↗ | not reported | — | — |
| MRCR v2 (128k average) | 60.1% ↗ | not reported | — | — |
| Only GLM-5.3 reports · 20 | ||||
| Agents' Last Exam | not reported | 28.5% ↗ | — | — |
| Agents' Last Exam (ALE-CLI) | not reported | 28.5% ↗ | — | — |
| AutomationBench v1.0.6 | not reported | 48.2% ↗ | — | — |
| CyberGym | not reported | 84.5% ↗ | — | — |
| DeepSWE v1.1 | not reported | 66.9% ↗ | — | — |
| ExploitBench | not reported | 54.4% ↗ | — | — |
| ExploitGym 2h | not reported | 105 ↗ | — | — |
| ExploitGym 6h | not reported | 130 ↗ | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | not reported | 29.3% ↗epoch | — | — |
| FrontierSWE | not reported | 78.1% ↗ | — | — |
| GDPval-AA v2 | not reported | 1769 ↗ | — | — |
| HLE (with tools) | not reported | 62.5% ↗ | — | — |
| NL2Repo | not reported | 58% ↗ | — | — |
| PostTrainBench | not reported | 39.8% ↗ | — | — |
| ProgramBench (Almost Solved) | not reported | 19% ↗ | — | — |
| SWE-Marathon v1.1 | not reported | 42.5% ↗ | — | — |
| Terminal-Bench 2.1 | not reported | 88.2% ↗ | — | — |
| Terminal-Bench 3.0 | not reported | 28.3% ↗ | — | — |
| Toolathlon Verified | not reported | 73% ↗ | — | — |
| Z.ai Code Bench (max effort) | not reported | 34.5% ↗ | — | — |
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 3.1 Flash-Lite minus GLM-5.3 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 3.1 Flash-Lite | GLM-5.3 | Edge |
|---|---|---|---|
| Context window | 1.0M tokens | 1M tokens | Gemini 3.1 Flash-Lite |
| Max output | 66K tokens | 131K tokens | GLM-5.3 |
| Input price / 1M | $0.25 ↗ | $1.4 ↗ | Gemini 3.1 Flash-Lite |
| Output price / 1M | $1.5 ↗ | $4.4 ↗ | Gemini 3.1 Flash-Lite |
| Cached input / 1M | $0.025 ↗ | $0.26 ↗ | Gemini 3.1 Flash-Lite |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $1.75 | $5.8 | Gemini 3.1 Flash-Lite |
| Modalities | text · vision · audio | text | Gemini 3.1 Flash-Lite |
| Released | 2026-03-03 | 2026-08-14 | — |
| Cited benchmark scores | 10 | 33 | — |
Reliability
Provider status
Last 60 days · refreshed daily on this page · live on /statusMore matchups
Gemini 3.1 Flash-Lite vs …
Models sharing the most benchmarksMore matchups
GLM-5.3 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Gemini 3.1 Flash-Lite and GLM-5.3 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.