DeepSeek V4 Pro vs GPT-5.6 Terra
DeepSeek V4 Pro 
GPT-5.6 Terra Verdict
- DeepSeek V4 Pro (max) and GPT-5.6 Terra (max) are level on quality (64.2 vs 65.0).
- DeepSeek V4 Pro (max) is stronger in agents & tools, composite, instruction following.
- GPT-5.6 Terra (max) is stronger in coding, knowledge, long context, maths, reasoning, multimodal.
- DeepSeek V4 Pro (max) is 8.3× cheaper ($0.544 vs $4.50 per 1M blended).
| Metric | DeepSeek V4 Pro (max) | GPT-5.6 Terra (max) |
|---|---|---|
| BenchLeader Index | 64.2 | 65.0 |
| Agents & tools score | 70.2 | 59.0 |
| Coding score | 58.8 | 67.2 |
| Composite score | 75.0 | 71.5 |
| Instruction following score | 74.3 | 69.8 |
| Knowledge score | 58.1 | 58.8 |
| Long context score | 66.6 | 68.0 |
| Maths score | 53.7 | 71.5 |
| Reasoning score | 70.9 | 72.3 |
| Multimodal score | – | 65.3 |
| Blended price $/M | $0.544 | $4.50 |
| Output speed | 81 tok/s | 84 tok/s |
| Time to first answer | 55.6 s | 220.9 s |
| Context window | 1M | 1.1M |
| GPQA Diamond | 89.7% | 93.3% |
| FrontierMath Tiers 1–3 | 45.3% | 86.0% |
| FrontierMath Tier 4 | 2.4% | 70.7% |
| OTIS Mock AIME | 96.7% | 99.7% |
| SWE-bench Verified (Epoch) | 77.6% | – |
| SimpleQA Verified | 47.0% | 43.2% |
| Terminal-Bench | – | 21.5% |
| SciCode | 50.0% | 53.9% |
| WeirdML | 48.9% | – |
| ProofBench | 16.0% | – |
| LiveBench | – | 77.9% |
| LiveBench Reasoning | – | 90.6% |
| LiveBench Coding | – | 78.3% |
| LiveBench Agentic Coding | – | 55.0% |
| LiveBench Mathematics | – | 94.9% |
| LiveBench Data Analysis | – | 79.3% |
| LiveBench Language | – | 82.9% |
| LiveBench Instruction Following | – | 64.6% |
| AA Intelligence Index | 36.3 | 42.3 |
| IFBench | 76.5% | 71.2% |
| AA-LCR | 80.3% | 83.0% |
| MMMU-Pro | – | 80.7% |
| AA-Omniscience | 0.8 | 0.1 |
| Terminal-Bench Hard | 46.2% | 57.6% |
| GPQA Diamond (AA) | 92.8% | 92.5% |
| Humanity's Last Exam (AA) | 41.0% | 42.9% |
| SciCode (AA) | 51.0% | 55.0% |
| τ²-Bench Telecom (AA) | 96.2% | 86.3% |
| LiveCodeBench | 87.5% | – |
| MMLU-Pro | 87.3% | – |
| IOI | – | 87.6% |
| LegalBench | 80.3% | 85.1% |
| CorpFin | 61.4% | – |
| TaxEval | 72.1% | – |
| Terminal-Bench 2.1 (Vals) | – | 77.5% |
| SWE-bench (Vals) | 77.4% | 95.4% |
| GPQA Diamond (Vals) | 89.4% | – |
| Vals Index | 42.9 | 59.6 |
| AIME 2026 | 96.7% | – |
| HMMT February 2026 | 93.9% | – |
| MathArena Apex | 28.1% | – |
| ARC-AGI-1 | – | 96.5% |
| ARC-AGI-2 | – | 83.9% |
| ARC-AGI-3 | – | 0.8% |
| CritPt | 18.0% | 30.0% |
| GDPval (AA) | 49.7% | 48.9% |
| τ²-Bench Banking (AA) | 39.6% | 40.2% |
| ITBench SRE (AA) | 38.3% | 51.0% |
| Analyst Agent (AA) | 18.8% | – |
| APEX-Agents (AA) | 24.3% | 38.9% |
| CaseLaw v2 | 59.4% | – |
| Code Migration | 26.2% | 47.8% |
| Excel Modeling Benchmark | 51.6% | 66.2% |
| Finance Agent v2 | 44.1% | 54.4% |
| Harvey's Legal Agent Benchmark | 3.8% | 0.8% |
| Legal Research Bench | 23.1% | 41.4% |
| MedCode | 40.5% | – |
| MedScribe | 75.1% | – |
| MMMU-Pro (Vals) | – | 86.5% |
| ProgramBench | 0.0% | 0.5% |
| Public Benefits Bench | 62.9% | – |
| SAGE | – | 47.0% |
| SkillsBench | 51.3% | 58.9% |
| Tax Agent Bench | – | 65.2% |
| Terminal-Bench 4.0 (Vals) | – | 26.3% |
| Terminal-Bench Science | – | 10.0% |
| Vibe Code Bench 1-100 | – | 14.8% |
| Vibe Code Bench v1.1 | 49.9% | 74.6% |
| Chess Puzzles | 20.0% | 54.0% |
| Mystery Game Puzzles | – | 35.0% |
| DeepSWE | – | 69.6% |
| LMCA | 41.2% | 52.2% |
| DTBench | 90.7% | 92.5% |
| CursorBench | – | 64.9% |
| ALE-Bench | – | 1951.4 |
Data as of 2026-09-19. Best configuration of each model; every score links to its source on the model pages.
DeepSeek V4 Pro vs GPT-5.6 Terra: questions
- Is DeepSeek V4 Pro better than GPT-5.6 Terra?
- DeepSeek V4 Pro (max) and GPT-5.6 Terra (max) are level on quality (64.2 vs 65.0). The BenchLeader Index combines every independent quality benchmark; GPT-5.6 Terra (max) is ahead overall as of 2026-09-19, but check the category scores for your use.
- Is DeepSeek V4 Pro better than GPT-5.6 Terra for coding?
- GPT-5.6 Terra scores higher in coding (67 vs 59 on the category index, where 50 is average).
- Is DeepSeek V4 Pro better than GPT-5.6 Terra for agentic tasks?
- DeepSeek V4 Pro scores higher in agentic tasks (70 vs 59 on the category index, where 50 is average).
- Which is cheaper, DeepSeek V4 Pro or GPT-5.6 Terra?
- DeepSeek V4 Pro is cheaper: $0.544 against $4.50 per million tokens, blended at three input tokens per output token.
- Which is faster, DeepSeek V4 Pro or GPT-5.6 Terra?
- GPT-5.6 Terra streams faster: 84 against 81 output tokens per second.
- Which has the larger context window?
- GPT-5.6 Terra accepts more context: 1.1M against 1M tokens.