Compare / head-to-head
DeepSeek V3.1vs
GLM-5.3
GLM-5.3 leads 3 of 3 shared benchmarks. GLM-5.3 has the larger context window (1M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
0 – 3
GLM-5.3 leads
Cheaper per token
—
list price, input + output
Larger context
GLM-5.3
1M tokens
Providers
2 providers
DeepSeek · Z.ai
DeepSeek V3.1
DeepSeek · released 2025-08-21
text
- Context
- 128K
- Max out
- —
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 18 · 9 core
Quality
Benchmark matrix
3 shared · 15 only DeepSeek V3.1 · 40 only GLM-5.3| Benchmark | DeepSeek V3.1 | GLM-5.3 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 3 | ||||
| LMArena Elo | 1417.1 ↗ | 1478.5 ↗ | -61.4 | GLM-5.3 |
| SimpleBench (SimpleBench) | 40% ↗ | 66.2% ↗ | -26.2 pt | GLM-5.3 |
| WeirdML (Håvard Tveit Ihle) | 38.4% ↗ | 75.4% ↗ | -37 pt | GLM-5.3 |
| Only DeepSeek V3.1 reports · 15 | ||||
| Aider-Polyglot (thinking) | 76.3% ↗ | not reported | — | — |
| AIME 2024 (thinking) | 93.1% ↗ | not reported | — | — |
| AIME 2025 (thinking) | 88.4% ↗ | not reported | — | — |
| BrowseComp | 30% ↗ | not reported | — | — |
| GPQA-Diamond (thinking) | 80.1% ↗ | not reported | — | — |
| HMMT 2025 (thinking) | 84.2% ↗ | not reported | — | — |
| Humanity's Last Exam (Python + Search) | 29.8% ↗ | not reported | — | — |
| Humanity's Last Exam (thinking, text-only) | 15.9% ↗ | not reported | — | — |
| LiveCodeBench (2408-2505) (thinking) | 74.8% ↗ | not reported | — | — |
| MMLU-Pro (thinking) | 84.8% ↗ | not reported | — | — |
| MMLU-Redux (thinking) | 93.7% ↗ | not reported | — | — |
| SimpleQA (search agent) | 93.4% ↗ | not reported | — | — |
| SWE Verified (Agent mode) | 66% ↗ | not reported | — | — |
| SWE-bench Multilingual (Agent mode) | 54.5% ↗ | not reported | — | — |
| Terminal-bench (Terminus 1 framework) | 31.3% ↗ | not reported | — | — |
| Only GLM-5.3 reports · 40 | ||||
| GPQA Diamond | not reported | 90.9% ↗epoch run | — | — |
| Agents' Last Exam | not reported | 28.5% ↗ | — | — |
| Agents' Last Exam (ALE-CLI) | not reported | 28.5% ↗ | — | — |
| APEX-Agents (Mercor) | not reported | 56.6% ↗ | — | — |
| AutomationBench v1.0.6 | not reported | 48.2% ↗ | — | — |
| Chess Puzzles (Epoch AI run) | not reported | 21% ↗epoch run | — | — |
| CyberGym | not reported | 84.5% ↗ | — | — |
| DeepSWE v1.1 | not reported | 66.9% ↗ | — | — |
| DeepSWE v1.1 (Datacurve) | not reported | 69% ↗ | — | — |
| 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 run | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | not reported | 68.8% ↗epoch run | — | — |
| FrontierSWE | not reported | 78.1% ↗ | — | — |
| FrontierSWE V2 (Proximal Labs) | not reported | 30.2% ↗ | — | — |
| GDPval-AA v2 | not reported | 1769 ↗ | — | — |
| HLE (with tools) | not reported | 62.5% ↗ | — | — |
| LiveBench Agentic Coding (LiveBench) | not reported | 60.9% ↗ | — | — |
| LiveBench Coding (LiveBench) | not reported | 79% ↗ | — | — |
| LiveBench Data Analysis (LiveBench) | not reported | 70.2% ↗ | — | — |
| LiveBench Instruction Following (LiveBench) | not reported | 69.3% ↗ | — | — |
| LiveBench Language (LiveBench) | not reported | 79.9% ↗ | — | — |
| LiveBench Mathematics (LiveBench) | not reported | 87.9% ↗ | — | — |
| LiveBench Reasoning (LiveBench) | not reported | 85.8% ↗ | — | — |
| LMArena Agent (LMArena) | not reported | 0.0236 ↗ | — | — |
| LMArena WebDev (LMArena) | not reported | 1623 ↗ | — | — |
| Mystery Game Puzzles (Epoch AI run) | not reported | 33% ↗epoch run | — | — |
| NL2Repo | not reported | 58% ↗ | — | — |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | not reported | 91.1% ↗epoch run | — | — |
| PostTrainBench | not reported | 39.8% ↗ | — | — |
| ProgramBench (Almost Solved) | not reported | 19% ↗ | — | — |
| SimpleQA Verified | not reported | 41% ↗epoch run | — | — |
| SWE-Marathon v1.1 | not reported | 42.5% ↗ | — | — |
| Terminal-Bench 2.1 | not reported | 88.2% ↗ | — | — |
| Terminal-Bench 3.0 | not reported | 28.3% ↗ | — | — |
| Terminal-Bench 4.0 (Vals AI) | not reported | 38.9% ↗ | — | — |
| Toolathlon Verified | not reported | 73% ↗ | — | — |
| Vending-Bench 2 (Andon Labs) | not reported | 8163.61 ↗ | — | — |
| 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 DeepSeek V3.1 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 | DeepSeek V3.1 | GLM-5.3 | Edge |
|---|---|---|---|
| Context window | 128K tokens | 1M tokens | GLM-5.3 |
| Max output | — | 131K tokens | — |
| Input price / 1M | — | $1.4 ↗ | — |
| Output price / 1M | — | $4.4 ↗ | — |
| Cached input / 1M | — | $0.26 ↗ | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | — | $5.8 | — |
| Modalities | text | text | Tie |
| Released | 2025-08-21 | 2026-08-18 | — |
| Cited benchmark scores | 18 | 57 | — |
Reliability
Provider status
Live from /statusMore matchups
DeepSeek V3.1 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 DeepSeek V3.1 and GLM-5.3 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.