DeepSeek V4.1 Flash 1017 vs GPT-5.6 Luna
DeepSeek V4.1 Flash 1017 leads GPT-5.6 Luna by 6.2 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. DeepSeek V4.1 Flash 1017 wins 11 and GPT-5.6 Luna wins 2 of 13 shared benchmarks. DeepSeek V4.1 Flash 1017 is cheaper for agent workloads.
Add more models to this comparison →Pillars
| Pillar | DeepSeek V4.1 Flash 1017 | GPT-5.6 Luna |
|---|---|---|
| Agentic coding | 63 | 59 |
| Long-horizon autonomy | 66 | 61 |
| Knowledge-work agents | 49 | 39 |
| Computer use | 45 | 40 |
| Security & context learning | 54 | 51 |
Shared benchmarks
| Benchmark | DeepSeek V4.1 Flash 1017 | GPT-5.6 Luna |
|---|---|---|
| APEX-Agentsindex | 69.8% | 43.0% |
| CursorBenchindex | 47.7% | 35.9% |
| ARC-AGI | 94.5% | 88.0% |
| ARC-AGI-2 | 78.1% | 59.5% |
| Chess Puzzles | 46.0% | 43.0% |
| CritPt | 24.0% | 20.6% |
| FrontierMath Tier 4 | 46.3% | 61.0% |
| FrontierMath Tiers 1-3 | 71.2% | 82.1% |
| GPQA Diamond | 93.2% | 91.6% |
| OTIS Mock AIME | 99.4% | 98.3% |
| SciCode | 55.9% | 53.6% |
| SimpleQA Verified | 42.5% | 41.0% |
| WeirdML v3 | 9.1% | 5.2% |
Specs and price
| DeepSeek V4.1 Flash 1017 | GPT-5.6 Luna | |
|---|---|---|
| Agent Index | 57.6 (#17) | 51.4 (#34) |
| Agent $/M | $0.050 | $0.10 |
| Input / output $/M | $0.15 / $0.60 | $0.20 / $1.20 |
| Context | 1M | 1.1M |
| Released | Oct 17, 2026 | Jul 9, 2026 |
| Weights | Open | Closed |
| API hosts | 51 | 5 |
Don’t pick one. Route.
Shogo can run DeepSeek V4.1 Flash 1017 and GPT-5.6 Luna side by side and send each step of an agent to the one that’s cheaper for it.
Try both in Shogo