Up to four models
GPT-5.6 Terra vs DeepSeek V4.1 Flash 1017
On the index (GPT-4 = 100), GPT-5.6 Terra scores 178 and DeepSeek V4.1 Flash 1017 176. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
GPT-5.6 Terra
178162–195
WeightsClosed
ReleasedJul 2026
Price in / out$2.0 / $12
Context1.05M
Core tests taken6 of 9
DeepSeek V4.1 Flash 1017176160–192
WeightsOpen
ReleasedOct 2026
Price in / out$0.30 / $1.2
Context1.05M
Core tests taken5 of 9
GPT-5.6 Terra and DeepSeek V4.1 Flash 1017 are too close to call (2 points apart, within the uncertainty), ahead on 3 of 5 shared core tests, at 8.6× the price.
Test by testthird-party results
GPT-5.6 TerraDeepSeek V4.1 Flash 1017
DeepSeek V4.1 Flash 1017Every result as a table
| Test | GPT-5.6 Terra | DeepSeek V4.1 Flash 1017 |
|---|---|---|
| Reasoning & knowledge | ||
| GPQA Diamond (Epoch AI) | 93.3% | 93.2% |
| SimpleQA Verified | 43.2% | 42.5% |
| Maths | ||
| FrontierMath (tiers 1-3) | 86.0% | 71.2% |
| AIME-style maths (OTIS mock) | 99.7% | 99.4% |
| Chess puzzles | 47.0% | 46.0% |
| FrontierMath (tier 4) | 70.7% | 46.3% |
| Coding | ||
| Terminal-Bench 4.0 | 22.7% | 31.8% |
| Code Migration | 47.8% | 44.2% |
| Vibe Code Bench v1.1 | 74.6% | 87.7% |
| Agents | ||
| Mystery game puzzles | 35.0% | 49.0% |
| Professional work | ||
| Finance Agent (v2) | 54.4% | 57.0% |
| EMB | 66.2% | 61.8% |
| Legal Research Bench | 41.3% | 45.2% |
| Tax Agent Bench | 30.0% | 44.4% |
| Novel problems | ||
| ARC-AGI-2 | 83.9% | 78.1% |
| ARC-AGI-1 | 96.5% | 94.5% |
Tests at least two of these models have taken. Bold rows are the core tests.
