Loading benchmark catalogue
Compare performance benchmarks, coding intelligence, reasoning evaluations, context limits, and API economics between GLM 5.2 (Zhipu AI) and DeepSeek V4.1 Flash 1017 (DeepSeek).
| Metric / Capability | GLM 5.2 | DeepSeek V4.1 Flash 1017 |
|---|---|---|
| Developer / Organization | Zhipu AI | DeepSeek |
| Release Date | 2026-06-16 | 2026-10-17 |
| Deployment / Weights | Open Weights | Open Weights |
| Context Window | 1.0M tokens (1,048,576 tokens) | 1.0M tokens (1,048,576 tokens) |
| Input Pricing (/1M Tokens) | Free (Open Weights) | Free (Open Weights) |
| Output Pricing (/1M Tokens) | Free (Open Weights) | Free (Open Weights) |
| Benchmark / Domain | GLM 5.2 | DeepSeek V4.1 Flash 1017 | Advantage |
|---|---|---|---|
Artificial Analysis Intelligence Index Composite index aggregating 10+ evaluations of general reasoning and knowledge. | 33.7% | 39.5% | DeepSeek V4.1 Flash 1017 (+17%) |
Artificial Analysis Coding Index Composite index of coding and software-engineering evaluations. | 68.8% | — | — |
Artificial Analysis Agentic Index Composite index of tool-use and long-horizon agentic tasks. | 38.4% | — | — |
Artificial Analysis Coding Agent Index v2.0 Composite index of coding-agent systems (model + harness) across FrontierCode v1.1, SWE-Atlas-QnA, and Terminal-Bench v4. | — | 46.3% | — |
SWE-bench Pro Harder, contamination-resistant successor to SWE-bench spanning multiple languages. | 62.1% | — | — |
Terminal-Bench 4.0 Sandboxed command-line tasks requiring shell and file-system operation. | — | 42.3% | — |
ARC-AGI-2 Novel visual grid puzzles requiring abstract rule induction. | 22.8% | — | — |
GPQA Diamond Graduate-level physics, chemistry, and biology questions written by domain experts. | 91.2% | 90.9% | GLM 5.2 leads |
Humanity's Last Exam Expert-written questions across hundreds of disciplines designed to resist saturation. | 40.5% | 42.6% | DeepSeek V4.1 Flash 1017 (+5%) |
FrontierMath v2 (Tiers 1-3) Unpublished, research-level mathematics problems across Tiers 1-3. | 59.2% | 67.4% | DeepSeek V4.1 Flash 1017 (+14%) |
Design Arena ELO Human-preference ELO across UI, 3D, data-visualization, and game-dev generations. | 1291 | 1362 | DeepSeek V4.1 Flash 1017 (+5%) |
DeepSeek V4.1 Flash 1017 outperforms GLM 5.2 across 4 evaluated benchmarks compared to 1 for GLM 5.2, excelling in Artificial Analysis Intelligence Index, Humanity's Last Exam, FrontierMath v2 (Tiers 1-3).
Both models feature equivalent context windows of 1.0M tokens (1,048,576 tokens).
Both models have comparable pricing structures or open weight deployment options.
GLM 5.2 is an open-weights model. DeepSeek V4.1 Flash 1017 is an open-weights model.
Analyze 130+ models side-by-side using interactive radar charts, cost-efficiency quadrant plots, and live dataset filters on the full NeuBench index.
Open Interactive Comparator