Model comparison
DeepSeek V4 Pro vs GPT-5.5 Instant
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 42.7 on the Noometry Index.
Last verified . 25 shared benchmarks.
DeepSeekSummary
- They share 25 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 8 categories and GPT-5.5 Instant in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek V4 Pro leads 64.8 to 26.5.
- The biggest single-benchmark swing is FrontierMath (Tiers 1-3): 64.6% for DeepSeek V4 Pro and 26.3% for GPT-5.5 Instant.
- DeepSeek V4 Pro has downloadable open weights; the other is API-only.
Side by side
| DeepSeek V4 Pro | GPT-5.5 Instant | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 54.3 | 42.7 |
| Released | 2026-04-24 | 2026-05-05 |
| Weights | Open | Proprietary |
| Context window | 1M | — |
| Max output | 393K | — |
| Input $ / M tokens | $0.66 | — |
| Output $ / M tokens | $1.98 | — |
| Results tracked | 48 | 27 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Pro leads
DeepSeek V4 Pro: 52.4 (#34), GPT-5.5 Instant: 44.3 (#74)
| Benchmark | DeepSeek V4 Pro | GPT-5.5 Instant |
|---|---|---|
| SciCode | 51% | 48.6% |
| LMArena Coding | 1470 | 1433 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| LMArena WebDev | 1582 | — |
| WeirdML | 66.2% | — |
| ALE-Bench | 1,403 | — |
Agentic & Tool Use Not comparable
DeepSeek V4 Pro: 32.8 (#58), GPT-5.5 Instant: —
| Benchmark | DeepSeek V4 Pro | GPT-5.5 Instant |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Vending-Bench 2 | 3,285 | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), GPT-5.5 Instant: 24.9 (#155)
| Benchmark | DeepSeek V4 Pro | GPT-5.5 Instant |
|---|---|---|
| CritPt | 18% | 0% |
| Chess Puzzles | 47% | 12% |
| LMArena Hard Prompts | 1461 | 1426 |
| Epoch Capabilities Index | 155.31 | 142.52 |
| ARC-AGI-2 | 61.3% | — |
| Kagi LLM Benchmark | 53.5% | — |
| NYT Connections (extended) | 91.3% | — |
| ARC-AGI-1 | 90.5% | — |
| Mystery Game Puzzles | 43% | — |
| DTBench | 93.9% | — |
| LMCA | 45.5% | — |
| Surface Evolver Bench | 40% | — |
| ForecastBench | 56.1 | — |
Math DeepSeek V4 Pro leads
DeepSeek V4 Pro: 64.8 (#30), GPT-5.5 Instant: 26.5 (#259)
| Benchmark | DeepSeek V4 Pro | GPT-5.5 Instant |
|---|---|---|
| FrontierMath (Tiers 1-3) | 64.6% | 26.3% |
| FrontierMath Tier 4 | 26.8% | 2.4% |
| OTIS Mock AIME 2024-2025 | 98.6% | 68.1% |
| LMArena Math | 1455 | 1420 |
| MathArena Final-Answer Competitions | 76.6% | — |
| ProofBench | 50% | — |
Knowledge DeepSeek V4 Pro leads
DeepSeek V4 Pro: 59.5 (#31), GPT-5.5 Instant: 48.9 (#74)
| Benchmark | DeepSeek V4 Pro | GPT-5.5 Instant |
|---|---|---|
| GPQA Diamond | 91.7% | 82.5% |
| LMArena Expert | 1464 | 1409 |
| SimpleQA Verified | 52.9% | — |
| Vectara Hallucination Rate | 8.6% | — |
Multimodal Not comparable
DeepSeek V4 Pro: —, GPT-5.5 Instant: 40.0 (#52)
| Benchmark | DeepSeek V4 Pro | GPT-5.5 Instant |
|---|---|---|
| LMArena Vision | — | 1250 |
| LMArena Document | — | 1403 |
Multilingual DeepSeek V4 Pro leads
DeepSeek V4 Pro: 54.4 (#45), GPT-5.5 Instant: 52.8 (#80)
| Benchmark | DeepSeek V4 Pro | GPT-5.5 Instant |
|---|---|---|
| LMArena Non-English | 1439 | 1417 |
| LMArena Chinese | 1486 | 1456 |
| LMArena French | 1472 | 1428 |
| LMArena German | 1458 | 1411 |
| LMArena Japanese | 1445 | 1408 |
| LMArena Korean | 1447 | 1392 |
| LMArena Russian | 1453 | 1431 |
| LMArena Spanish | 1458 | 1429 |
Instruction Following DeepSeek V4 Pro leads
DeepSeek V4 Pro: 76.1 (#47), GPT-5.5 Instant: 74.2 (#100)
| Benchmark | DeepSeek V4 Pro | GPT-5.5 Instant |
|---|---|---|
| LMArena Instruction Following | 1448 | 1406 |
Long Context DeepSeek V4 Pro leads
DeepSeek V4 Pro: 45.0 (#51), GPT-5.5 Instant: 43.4 (#96)
| Benchmark | DeepSeek V4 Pro | GPT-5.5 Instant |
|---|---|---|
| LMArena Longer Query | 1458 | 1422 |
| CL-bench Life | 13.5% | — |
Writing & Preference DeepSeek V4 Pro leads
DeepSeek V4 Pro: 65.5 (#46), GPT-5.5 Instant: 61.8 (#85)
| Benchmark | DeepSeek V4 Pro | GPT-5.5 Instant |
|---|---|---|
| LMArena Text | 1451 | 1419 |
| LMArena Creative Writing | 1446 | 1419 |
| LMArena Multi-Turn | 1467 | 1433 |
| EQ-Bench Creative Writing | 1553 | — |
| EQ-Bench 4 | 1166 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than GPT-5.5 Instant?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 42.7 on the Noometry Index.
Is DeepSeek V4 Pro or GPT-5.5 Instant better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 44.3 in the Noometry coding category.
How many benchmarks do DeepSeek V4 Pro and GPT-5.5 Instant share?
25 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and GPT-5.5 Instant has 27.