DeepSeek V4.1 Flash 1017 vs GLM 5.2
Full profiles: DeepSeek V4.1 Flash 1017 (DeepSeek) and GLM 5.2 (Zhipu AI): every benchmark, price and rank.
Updated · Data from 7 sources
As of 25 Oct 2026, DeepSeek V4.1 Flash 1017 (max) and GLM 5.2 (max) are neck and neck on quality (64.0 vs 62.3 on the BenchLeader Index, inside the ±5.3 margin of error), and DeepSeek V4.1 Flash 1017 (max) costs 4.1× less, which makes it the better buy for most workloads. DeepSeek V4.1 Flash 1017 leads in every category we measure both in.
DeepSeek V4.1 Flash 1017 benchmarks max DeepSeek
GLM 5.2 benchmarks max Zhipu AIOpen in the comparison toolWhich to pick for…
The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.
| Use case | Pick | Why | Confidence |
|---|---|---|---|
| Coding | DeepSeek V4.1 Flash 1017 | 70 vs 62 category score; Arena WebDev 1665 vs 1604 | High |
| Agents & tool use | Either | 64 vs 63 category score; Harvey's Legal Agent Benchmark v1 9.2% vs 7.1% | Either |
| Maths | DeepSeek V4.1 Flash 1017 | 62 vs 57 category score; FrontierMath Tiers 1–3 71.2% vs 59.2% | Medium |
| Reasoning | Either | 71 vs 70 category score; GPQA Diamond 93.2% vs 91.9% | Either |
| Knowledge | DeepSeek V4.1 Flash 1017 | 57 vs 54 category score; SimpleQA Verified 42.5% vs 34.2% | Medium |
| Long documents | DeepSeek V4.1 Flash 1017 | 66 vs 64 category score; AA-LCR 82.7% vs 78.3% | Medium |
| Chat quality (human preference) | – | Only GLM 5.2 has an Arena rating (67 category score) | Not comparable |
| Lowest cost | DeepSeek V4.1 Flash 1017 | $0.525 vs $2.15 per 1M blended; agentic coding session $0.023 vs $0.102 | – |
| Fastest first answer | DeepSeek V4.1 Flash 1017 | 11.4 s (DeepSeek V4.1 Flash 1017 (max)) vs 24.4 s (GLM 5.2 (max)) to the first answer token | – |
| High volume on a budget | Neither flagship | A smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs | – |
Key differences
- DeepSeek V4.1 Flash 1017 (max) and GLM 5.2 (max) are neck and neck on quality (64.0 vs 62.3, within the ±5.3 margin).
- DeepSeek V4.1 Flash 1017 (max) is stronger in agents & tools, coding, composite, knowledge, long context, maths, multimodal, reasoning.
- GLM 5.2 (max) is stronger in human preference, instruction following.
- DeepSeek V4.1 Flash 1017 (max) is 4.1× cheaper ($0.525 vs $2.15 per 1M blended).
- DeepSeek V4.1 Flash 1017 (max) streams 2.1× faster (195 vs 94 tokens per second).
At a glance
| Spec | DeepSeek V4.1 Flash 1017 | GLM 5.2 |
|---|---|---|
| Developer | DeepSeek | Zhipu AI |
| Released | 17 Oct 2026 | 16 Jun 2026 |
| Context window (tokens) | 1,048,576 | 1,000,000 |
| Compared at | DeepSeek V4.1 Flash 1017 (max) | GLM 5.2 (max) |
| Open weights | Yes | Yes |
| Price in / out / cached input per 1M | $0.300 / $1.20 / $0.0060 | $1.40 / $4.40 / $0.299 |
Category by category
Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.
| Category | DeepSeek V4.1 Flash 1017 (max) | GLM 5.2 (max) | Leader |
|---|---|---|---|
| Reasoning | 70.6 | 69.7 | level |
| Coding | 69.9 | 62.2 | DeepSeek V4.1 Flash 1017 |
| Agents & tools | 63.5 | 62.6 | level |
| Maths | 61.8 | 57.3 | DeepSeek V4.1 Flash 1017 |
| Knowledge | 57.1 | 53.7 | DeepSeek V4.1 Flash 1017 |
| Instruction following | – | 71.7 | not comparable |
| Human preference | – | 66.6 | not comparable |
| Multimodal | 63.7 | – | not comparable |
| Long context | 66.4 | 64.2 | DeepSeek V4.1 Flash 1017 |
| Composite | 80.0 | 67.2 | DeepSeek V4.1 Flash 1017 |
Where sources disagree
Across the 24 benchmarks both have, DeepSeek V4.1 Flash 1017 leads 22, GLM 5.2 leads 1 and 1 is a tie. Every overall index that rates both puts DeepSeek V4.1 Flash 1017 ahead.
| Overall index | DeepSeek V4.1 Flash 1017 (max) | GLM 5.2 (max) | Leader |
|---|---|---|---|
| BenchLeader Index | 64.0 | 62.3 | neck and neck |
| AA Intelligence Index v4.3.2 | 45.1 | 33.7 | DeepSeek V4.1 Flash 1017 |
Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).
Price, speed and latency
List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page; suite cost is what Artificial Analysis paid to run its whole Intelligence Index.
| Measure | DeepSeek V4.1 Flash 1017 (max) | GLM 5.2 (max) |
|---|---|---|
| Blended price per 1M tokens | $0.525 | $2.15 |
| Input / output per 1M | $0.300 / $1.20 | $1.40 / $4.40 |
| Cached input per 1M | $0.0060 | $0.299 |
| Price per task: chat reply | $0.0005 | $0.0019 |
| Price per task: summarize a 30-page report | $0.0043 | $0.019 |
| Price per task: code edit | $0.0036 | $0.015 |
| Price per task: agentic coding session | $0.023 | $0.102 |
| Price per task: structured extraction | $0.0008 | $0.0037 |
| Cost to run a full benchmark suite | $0.262 | $1.47 |
| Output speed | 195 tok/s | 94 tok/s |
| Time to first answer | 11.4 s | 24.4 s |
| Time to a complete answer | 13.9 s | 29.7 s |
Per token is not per task: on Artificial Analysis' full suite DeepSeek V4.1 Flash 1017 (max) cost $0.262 and GLM 5.2 (max) $1.47, because GLM 5.2 (max) used about 1.4× as many tokens for the same work, which a per-token price hides.
Each model is compared at its best-ranked reasoning-effort setting. DeepSeek V4.1 Flash 1017 (max) scores 64.0 and takes 11 s to its first answer token. GLM 5.2 (max) scores 62.3 and takes 24 s to its first answer token; GLM 5.2 scores 51.8 in 0 s, GLM 5.2 (no reasoning) scores 48.0 in 1 s. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.
Every benchmark
124 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.
| Metric | Publisher | DeepSeek V4.1 Flash 1017 (max) | GLM 5.2 (max) |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 64.0 | 62.3 |
| Agents & tools score | BenchLeader | 63.5 | 62.6 |
| Coding score | BenchLeader | 69.9 | 62.2 |
| Composite score | BenchLeader | 80.0 | 67.2 |
| Knowledge score | BenchLeader | 57.1 | 53.7 |
| Long context score | BenchLeader | 66.4 | 64.2 |
| Maths score | BenchLeader | 61.8 | 57.3 |
| Multimodal score | BenchLeader | 63.7 | – |
| Reasoning score | BenchLeader | 70.6 | 69.7 |
| Human preference score | BenchLeader | – | 66.6 |
| Instruction following score | BenchLeader | – | 71.7 |
| Blended price $/M | BenchLeader | $0.525 | $2.15 |
| Output speed | BenchLeader | 195 tok/s | 94 tok/s |
| Time to first answer | BenchLeader | 11.4 s | 24.4 s |
| Context window | BenchLeader | 1,048,576 tokens | 1,000,000 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | 93.2% | 91.9% |
| SimpleBench | SimpleBench | – | 58.8%† |
| Arena Hard Prompts | Arena (LMArena) | – | 1496 |
| LiveBench 2.0 Reasoning | LiveBench | 89.1%† | – |
| LiveBench 1.0 Reasoning | LiveBench | – | 78.6%† |
| GPQA Diamond (AA) | Artificial Analysis | – | 89.5% |
| Humanity's Last Exam (AA) | Artificial Analysis | 43.3% | 41.1% |
| GPQA Diamond (Vals) | Vals AI | – | 85.6%† |
| Kagi LLM Benchmark | Kagi LLM Benchmark | – | 62.6%† |
| ARC-AGI-1 | ARC Prize | 93.5% | 77.0%† |
| ARC-AGI-2 | ARC Prize | 78.1% | 22.8%† |
| CritPt | Artificial Analysis | 24.0% | 20.9% |
| Chess Puzzles | Epoch AI Benchmarking Hub | 46.0% | 22.0% |
| Mystery Game Puzzles | Epoch AI Benchmarking Hub | 49.0% | 18.0%† |
| LMCA | Epoch AI Benchmarking Hub | – | 45.8% |
| DTBench | Epoch AI Benchmarking Hub | – | 93.6% |
| Coding | |||
| SWE-bench Verified (Epoch) | Epoch AI Benchmarking Hub | – | 78.7% |
| SciCode | SciCode | – | 51.2% |
| WeirdML | WeirdML | – | 70.1% |
| FrontierCode | Cognition | – | 24.5%† |
| Arena Coding | Arena (LMArena) | – | 1514 |
| Arena WebDev | Arena (LMArena) | 1665 | 1604 |
| LiveBench 2.0 Coding | LiveBench | 96.0%† | – |
| LiveBench 1.0 Coding | LiveBench | – | 79.7%† |
| SciCode (AA) | Artificial Analysis | 55.9% | 51.2% |
Show all 124 benchmarks
| Metric | Publisher | DeepSeek V4.1 Flash 1017 (max) | GLM 5.2 (max) |
|---|---|---|---|
| Coding | |||
| LiveCodeBench | Vals AI | – | 69.5%† |
| SWE-bench (Vals) | Vals AI | – | 82.8% |
| Code Migration | Vals AI | 44.2% | 37.9% |
| Code Migration v1.1 | Vals AI | 67.5% | – |
| ProgramBench | Vals AI | – | 0.5% |
| Vibe Code Bench v1.1 | Vals AI | 87.7% | 64.0% |
| SWE Atlas: Codebase QnA | Scale AI SEAL | – | 48.1%† |
| SWE Atlas: Refactoring | Scale AI SEAL | – | 42.4%† |
| SWE Atlas: Test Writing | Scale AI SEAL | – | 41.5%† |
| Surface Evolver Bench | Epoch AI Benchmarking Hub | – | 55.6%† |
| DeepSWE v1.1 | Epoch AI Benchmarking Hub | – | 43.8% |
| ALE-Bench | Epoch AI Benchmarking Hub | – | 1010.2 |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 33.8% | 1.0% |
| Terminal-Bench 2.1 (AA) | Artificial Analysis | – | 77.9% |
| Startrise Frontend | Startrise | – | 72.1%† |
| FrontierCode 1.1 (AA) | Artificial Analysis | 43.7% | – |
| Agents & tools | |||
| APEX-Agents | Mercor | – | 45.2%† |
| Arena Agent | Arena (LMArena) | – | 1.7 |
| LiveBench Long Agentic | LiveBench | 50.9%† | – |
| LiveBench Browser | LiveBench | 76.0%† | – |
| LiveBench Agentic Coding | LiveBench | – | 51.8%† |
| Terminal-Bench Hard | Artificial Analysis | – | 50.8% |
| τ²-Bench Telecom (AA) | Artificial Analysis | – | 99.1% |
| Terminal-Bench 2.1 (Vals) | Vals AI | – | 67.8% |
| MCP Atlas | Scale AI SEAL | – | 77.8%† |
| HiL-Bench | Scale AI SEAL | – | 43.7%† |
| τ³-bench Banking | τ²-bench | – | 37.1%† |
| GDPval-AA v2.1 | Artificial Analysis | 62.3% | – |
| τ³-Banking (AA) | Artificial Analysis | – | 34.6% |
| ITBench SRE (AA) | Artificial Analysis | – | 42.7% |
| APEX-Agents (AA) | Artificial Analysis | – | 33.7% |
| Finance Agent v2 | Vals AI | 57.0% | 49.7%† |
| Finance Agent v2.1 | Vals AI | 62.9% | – |
| Harvey's Legal Agent Benchmark v1.1 | Vals AI | 14.2% | – |
| Harvey's Legal Agent Benchmark v1 | Vals AI | 9.2% | 7.1% |
| Legal Research Bench | Vals AI | 45.2% | 31.3% |
| SkillsBench | Vals AI | – | 45.1% |
| SREBench | Vals AI | – | 0.0% |
| Tax Agent Bench | Vals AI | 44.4% | – |
| Terminal-Bench 4.0 (Vals) | Vals AI | 31.8% | – |
| PostTrainBench | Epoch AI Benchmarking Hub | – | 31.7% |
| GBAEval | Epoch AI Benchmarking Hub | – | 0.0%† |
| Vending-Bench 2 | Andon Labs | – | 8313.8† |
| AutomationBench | Artificial Analysis | 66.6% | 28.4% |
| GDP.pdf (AA) | Artificial Analysis | 17.2% | 10.4% |
| EnterpriseOps-Gym | Artificial Analysis | – | 42.7% |
| AA-Briefcase v1.1 | Artificial Analysis | 1570 | 1227 |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | 12.9% | – |
| CWE-Bench (AA) | Artificial Analysis | 69.2% | – |
| DeepSec-Bench (AA) | Artificial Analysis | 24.9% | – |
| CyberGym end-to-end (AA) | Artificial Analysis | 76.3% | – |
| Maths | |||
| FrontierMath Tiers 1–3 | Epoch AI Benchmarking Hub | 71.2% | 59.2% |
| FrontierMath Tier 4 | Epoch AI Benchmarking Hub | 46.3% | 29.3% |
| OTIS Mock AIME | Epoch AI Benchmarking Hub | 99.4% | 86.4% |
| ProofBench | Vals AI | – | 35.0% |
| LiveBench Mathematics | LiveBench | – | 89.8%† |
| AIME 2026 | MathArena | – | 90.0%† |
| HMMT February 2026 | MathArena | – | 92.4%† |
| MathArena Apex | MathArena | – | 19.8%† |
| Arena Maths | Arena (LMArena) | – | 1488 |
| Knowledge | |||
| SimpleQA Verified | Epoch AI Benchmarking Hub | 42.5% | 34.2% |
| LiveBench 2.0 Data Analysis | LiveBench | 79.5%† | – |
| LiveBench Research | LiveBench | 79.5%† | – |
| LiveBench 1.0 Data Analysis | LiveBench | – | 73.7%† |
| AA-Omniscience | Artificial Analysis | 4.7 | 4.4 |
| MMLU-Pro | Vals AI | – | 86.7%† |
| LegalBench | Vals AI | – | 84.1%† |
| CorpFin | Vals AI | – | 66.1%† |
| TaxEval | Vals AI | – | 73.3%† |
| Excel Modeling Benchmark | Vals AI | 61.8% | 61.5%† |
| Excel Modeling Benchmark v1.1 | Vals AI | 69.2% | – |
| MedCode | Vals AI | – | 40.8%† |
| MedScribe | Vals AI | – | 83.5%† |
| DrugDiscoveryBench | Scale AI SEAL | – | 36.2%† |
| EBR-bench | Epoch AI Benchmarking Hub | – | 9.5% |
| AA-Omniscience: accuracy | Artificial Analysis | 42.9% | 24.3% |
| AA-Omniscience: non-hallucination | Artificial Analysis | 33.3% | 73.7% |
| Instruction following | |||
| LiveBench Language | LiveBench | – | 76.2%† |
| LiveBench Instruction Following | LiveBench | – | 62.3%† |
| IFBench | Artificial Analysis | – | 73.3% |
| Arena Instruction Following | Arena (LMArena) | – | 1469 |
| Human preference | |||
| Arena Text | Arena (LMArena) | – | 1475 |
| EQ-Bench 4 | EQ-Bench | – | 1211† |
| Arena Creative Writing | Arena (LMArena) | – | 1453 |
| Arena Multi-turn | Arena (LMArena) | – | 1475 |
| Arena Longer Queries | Arena (LMArena) | – | 1486 |
| Multimodal | |||
| LiveBench Design | LiveBench | 83.4%† | – |
| MMMU-Pro | Artificial Analysis | 80.6% | – |
| Long context | |||
| AA-LCR | Artificial Analysis | 82.7% | 78.3% |
| MLCR | Artificial Analysis | 26.7% | 12.8% |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | – | 151.9† |
| LiveBench 2.0 | LiveBench | 79.2%† | – |
| LiveBench 1.0 | LiveBench | – | 73.2%† |
| AA Intelligence Index v4.3.2 | Artificial Analysis | 45.1 | 33.7 |
| Vals Index | Vals AI | 61.3 | – |
| Vals Finance Index | Vals AI | 60.9% | – |
| Vals Router Index | Vals AI | 61.3%† | – |
| AA Cyber Index | Artificial Analysis | 56.8 | – |
| Metadata | |||
| AA Openness Index | Artificial Analysis | 44.4 | 44.4 |
Data as of 2026-10-25. Best configuration of each model; every score links to its source on the model pages.
What changed
- GLM 5.2 (max): 1 new result (AA-Briefcase v1.1); cached input price moved from $0.26/M to $0.30/M (15% up).
- GLM 5.2 (max): cached input price moved from $0.30/M to $0.26/M (13% down).
- GLM 5.2 (max): BenchLeader Index 62.8 → 62.3.
- DeepSeek V4.1 Flash 1017 (max): 10 new results (MLCR, AA Cyber Index, CWE-Bench (AA)…).
- GLM 5.2 (max): 1 new result (Harvey's Legal Agent Benchmark v1).
- DeepSeek V4.1 Flash 1017 (max): BenchLeader Index 63.1 → 63.9.
- DeepSeek V4.1 Flash 1017 (max): 11 new results (GPQA Diamond, FrontierMath Tiers 1–3, FrontierMath Tier 4…).
- GLM 5.2 (max): 1 score revision.
- DeepSeek V4.1 Flash 1017 (max): 1 new result (FrontierCode 1.1 (AA)); 2 score revisions.
- GLM 5.2 (max): cached input price moved from $0.26/M to $0.30/M (15% up).
Related comparisons
- DeepSeek V4.1 Flash 1017 vs Claude Opus 5.5
- DeepSeek V4.1 Flash 1017 vs Gemini 4 Argon
- DeepSeek V4.1 Flash 1017 vs Muse Wonder Full
- GLM 5.2 vs Claude Opus 5.5
- GLM 5.2 vs Gemini 4 Argon
- GLM 5.2 vs Muse Wonder Full
- GLM 5.2 vs Claude Fable 5.1
- DeepSeek V4.1 Flash 1017 vs GPT-6 Astra
Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs
DeepSeek V4.1 Flash 1017 vs GLM 5.2: questions
- Is DeepSeek V4.1 Flash 1017 better than GLM 5.2?
- DeepSeek V4.1 Flash 1017 (max) and GLM 5.2 (max) are neck and neck on quality (64.0 vs 62.3, within the ±5.3 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 2026-10-25, so check the category scores for your use.
- Which is cheaper, DeepSeek V4.1 Flash 1017 or GLM 5.2, and will that change?
- DeepSeek V4.1 Flash 1017 is cheaper today: $0.525 against $2.15 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
- Is DeepSeek V4.1 Flash 1017 better than GLM 5.2 for coding?
- DeepSeek V4.1 Flash 1017 scores higher in coding (70 vs 62 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1665 for DeepSeek V4.1 Flash 1017 against 1604 for GLM 5.2.
- Is DeepSeek V4.1 Flash 1017 better than GLM 5.2 for agents and tool use?
- They are level in agentic tasks (64 vs 63 on the category index).
- Which is better for long documents?
- DeepSeek V4.1 Flash 1017 scores higher on long-context benchmarks (66 vs 64). Both take about 1M tokens: DeepSeek V4.1 Flash 1017 1,048,576, GLM 5.2 1,000,000.
- Which has lower latency?
- DeepSeek V4.1 Flash 1017 (max) answers first: 11.4 s against 24.4 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
- Which is faster, DeepSeek V4.1 Flash 1017 or GLM 5.2?
- DeepSeek V4.1 Flash 1017 streams faster: 195 against 94 output tokens per second.
- Do independent benchmarks agree on DeepSeek V4.1 Flash 1017 vs GLM 5.2?
- Across the 24 benchmarks both have, DeepSeek V4.1 Flash 1017 leads 22, GLM 5.2 leads 1 and 1 is a tie. Every overall index that rates both puts DeepSeek V4.1 Flash 1017 ahead.
- When was DeepSeek V4.1 Flash 1017 released?
- DeepSeek V4.1 Flash 1017 was released on 17 Oct 2026.