Model comparison
DeepSeek-V3.2-Speciale vs Grok 4.1 Fast
Grok 4.1 Fast is the stronger model overall, scoring 41.4 to 39.7 on the Noometry Index.
Last verified . 2 shared benchmarks.
DeepSeekSummary
- They share 2 benchmarks with published results for both. DeepSeek-V3.2-Speciale scores higher in 1 category and Grok 4.1 Fast in 2 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Grok 4.1 Fast leads 57.2 to 46.0.
- Grok 4.1 Fast is cheaper at $0.20 / $0.50 per million input/output tokens, against $0.58 / $1.68 for DeepSeek-V3.2-Speciale.
- DeepSeek-V3.2-Speciale has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V3.2-Speciale | Grok 4.1 Fast | |
|---|---|---|
| Provider | DeepSeek | xAI |
| Noometry Index | 39.7 | 41.4 |
| Released | 2025-12-01 | 2025-06-27 |
| Weights | Open | Proprietary |
| Context window | 128K | 128K |
| Max output | 128K | 30K |
| Input $ / M tokens | $0.58 | $0.20 |
| Output $ / M tokens | $1.68 | $0.50 |
| Results tracked | 3 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.2-Speciale leads
DeepSeek-V3.2-Speciale: 40.4 (#140), Grok 4.1 Fast: 34.1 (#245)
| Benchmark | DeepSeek-V3.2-Speciale | Grok 4.1 Fast |
|---|---|---|
| LMArena WebDev | — | 1242 |
| WeirdML | 46.7% | — |
| LMArena Coding | — | 1411 |
| ALE-Bench | — | 394.93 |
Agentic & Tool Use Not comparable
DeepSeek-V3.2-Speciale: —, Grok 4.1 Fast: 36.3 (#39)
| Benchmark | DeepSeek-V3.2-Speciale | Grok 4.1 Fast |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 69.6% |
| τ²-bench Banking | — | 13.1% |
| LMArena Search | — | 1171 |
| Vending-Bench 2 | — | 1,107 |
Reasoning Grok 4.1 Fast leads
DeepSeek-V3.2-Speciale: 32.9 (#73), Grok 4.1 Fast: 43.4 (#49)
| Benchmark | DeepSeek-V3.2-Speciale | Grok 4.1 Fast |
|---|---|---|
| SimpleBench | 52.6% | 56% |
| NYT Connections (extended) | — | 87.4% |
| LMArena Hard Prompts | — | 1407 |
| DTBench | — | 87.7% |
| ForecastBench | — | 61 |
Math Not comparable
DeepSeek-V3.2-Speciale: —, Grok 4.1 Fast: 31.9 (#221)
| Benchmark | DeepSeek-V3.2-Speciale | Grok 4.1 Fast |
|---|---|---|
| MathArena Final-Answer Competitions | — | 60.9% |
| ProofBench | — | 4% |
| LMArena Math | — | 1408 |
Knowledge Not comparable
DeepSeek-V3.2-Speciale: —, Grok 4.1 Fast: 33.1 (#207)
| Benchmark | DeepSeek-V3.2-Speciale | Grok 4.1 Fast |
|---|---|---|
| Vectara Hallucination Rate | — | 17.8% |
| LMArena Expert | — | 1399 |
Multimodal Not comparable
DeepSeek-V3.2-Speciale: —, Grok 4.1 Fast: 37.0 (#76)
| Benchmark | DeepSeek-V3.2-Speciale | Grok 4.1 Fast |
|---|---|---|
| LMArena Vision | — | 1201 |
Multilingual Not comparable
DeepSeek-V3.2-Speciale: —, Grok 4.1 Fast: 51.0 (#114)
| Benchmark | DeepSeek-V3.2-Speciale | Grok 4.1 Fast |
|---|---|---|
| LMArena Non-English | — | 1391 |
| LMArena Chinese | — | 1441 |
| LMArena French | — | 1415 |
| LMArena German | — | 1404 |
| LMArena Japanese | — | 1349 |
| LMArena Korean | — | 1361 |
| LMArena Russian | — | 1387 |
| LMArena Spanish | — | 1413 |
Instruction Following Not comparable
DeepSeek-V3.2-Speciale: —, Grok 4.1 Fast: 72.7 (#133)
| Benchmark | DeepSeek-V3.2-Speciale | Grok 4.1 Fast |
|---|---|---|
| LMArena Instruction Following | — | 1376 |
Long Context Not comparable
DeepSeek-V3.2-Speciale: —, Grok 4.1 Fast: 42.4 (#126)
| Benchmark | DeepSeek-V3.2-Speciale | Grok 4.1 Fast |
|---|---|---|
| LMArena Longer Query | — | 1390 |
Writing & Preference Grok 4.1 Fast leads
DeepSeek-V3.2-Speciale: 46.0 (#222), Grok 4.1 Fast: 57.2 (#131)
| Benchmark | DeepSeek-V3.2-Speciale | Grok 4.1 Fast |
|---|---|---|
| EQ-Bench Creative Writing | 1276 | 1327 |
| LMArena Text | — | 1408 |
| LMArena Creative Writing | — | 1394 |
| LMArena Multi-Turn | — | 1389 |
Frequently asked questions
Is DeepSeek-V3.2-Speciale better than Grok 4.1 Fast?
Grok 4.1 Fast is the stronger model overall, scoring 41.4 to 39.7 on the Noometry Index.
Which is cheaper, DeepSeek-V3.2-Speciale or Grok 4.1 Fast?
Grok 4.1 Fast is cheaper. It lists at $0.20 per million input tokens and $0.50 per million output tokens; DeepSeek-V3.2-Speciale lists at $0.58 and $1.68.
Is DeepSeek-V3.2-Speciale or Grok 4.1 Fast better for coding?
DeepSeek-V3.2-Speciale scores higher on coding benchmarks: 40.4 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
Both accept 128K tokens.
How many benchmarks do DeepSeek-V3.2-Speciale and Grok 4.1 Fast share?
2 benchmarks have published results for both models. DeepSeek-V3.2-Speciale has 3 scored results on Noometry and Grok 4.1 Fast has 32.