Model comparison
DeepSeek-V2.5 (Sep 2024) vs Grok 4.3
Grok 4.3 is the stronger model overall, scoring 43.8 to 37.6 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeekSummary
- They share 17 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 0 categories and Grok 4.3 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Grok 4.3 leads 52.5 to 34.8.
- DeepSeek-V2.5 (Sep 2024) has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Grok 4.3 | |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 37.6 | 43.8 |
| Released | 2024-09-06 | 2026-04-17 |
| Weights | Open | Proprietary |
| Context window | — | 1M |
| Max output | — | 30K |
| Input $ / M tokens | — | $1.25 |
| Output $ / M tokens | — | $2.50 |
| Results tracked | 22 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.3 leads
DeepSeek-V2.5 (Sep 2024): 31.7 (#281), Grok 4.3: 41.6 (#121)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Grok 4.3 |
|---|---|---|
| LMArena Coding | 1309 | 1415 |
| Aider Polyglot | 17.8% | — |
| LMArena WebDev | — | 1357 |
| SciCode | — | 47.3% |
| WeirdML | — | 49.9% |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| ALE-Bench | — | 944.17 |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Agentic & Tool Use Not comparable
DeepSeek-V2.5 (Sep 2024): —, Grok 4.3: 27.7 (#99)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Grok 4.3 |
|---|---|---|
| GDP.pdf | — | 8% |
| LMArena Search | — | 1165 |
| Vending-Bench 2 | — | 35.26 |
Reasoning Grok 4.3 leads
DeepSeek-V2.5 (Sep 2024): 25.6 (#145), Grok 4.3: 35.9 (#68)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Grok 4.3 |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1396 |
| NYT Connections (extended) | — | 55.2% |
| CritPt | — | 8% |
| Chess Puzzles | — | 25% |
| DTBench | — | 90.7% |
| LMCA | — | 38.3% |
| Epoch Capabilities Index | — | 149.16 |
| ForecastBench | — | 60.3 |
Math Grok 4.3 leads
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), Grok 4.3: 46.0 (#74)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Grok 4.3 |
|---|---|---|
| LMArena Math | 1288 | 1388 |
| FrontierMath (Tiers 1-3) | — | 42.8% |
| FrontierMath Tier 4 | — | 14.6% |
| OTIS Mock AIME 2024-2025 | — | 93.3% |
| ProofBench | — | 11% |
Knowledge Grok 4.3 leads
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), Grok 4.3: 52.5 (#62)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Grok 4.3 |
|---|---|---|
| LMArena Expert | 1266 | 1385 |
| GPQA Diamond | — | 88.8% |
| SimpleQA Verified | — | 33.2% |
Multimodal Not comparable
DeepSeek-V2.5 (Sep 2024): —, Grok 4.3: 31.6 (#104)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Grok 4.3 |
|---|---|---|
| LMArena Vision | — | 1229 |
| Blueprint-Bench 2 | — | 0% |
Multilingual Grok 4.3 leads
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), Grok 4.3: 50.5 (#120)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Grok 4.3 |
|---|---|---|
| LMArena Non-English | 1273 | 1385 |
| LMArena Chinese | 1318 | 1422 |
| LMArena French | 1289 | 1412 |
| LMArena German | 1258 | 1395 |
| LMArena Japanese | 1228 | 1379 |
| LMArena Korean | 1209 | 1356 |
| LMArena Russian | 1289 | 1399 |
| LMArena Spanish | 1248 | 1398 |
Instruction Following Grok 4.3 leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), Grok 4.3: 72.1 (#140)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Grok 4.3 |
|---|---|---|
| LMArena Instruction Following | 1280 | 1366 |
Long Context Grok 4.3 leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), Grok 4.3: 42.5 (#123)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Grok 4.3 |
|---|---|---|
| LMArena Longer Query | 1301 | 1393 |
Writing & Preference Grok 4.3 leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), Grok 4.3: 58.5 (#118)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Grok 4.3 |
|---|---|---|
| LMArena Text | 1294 | 1397 |
| LMArena Creative Writing | 1285 | 1380 |
| LMArena Multi-Turn | 1297 | 1406 |
| EQ-Bench 4 | — | 1075 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Grok 4.3?
Grok 4.3 is the stronger model overall, scoring 43.8 to 37.6 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Grok 4.3 better for coding?
Grok 4.3 scores higher on coding benchmarks: 41.6 versus 31.7 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Grok 4.3 share?
17 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Grok 4.3 has 40.