Model comparison
DeepSeek-V3.2-Speciale vs Nova Premier 1.0
DeepSeek-V3.2-Speciale is the stronger model overall, scoring 39.1 to 37.7 on the Noometry Index.
Last verified . 0 shared benchmarks.
DeepSeekSummary
- The widest gap is in reasoning, where DeepSeek-V3.2-Speciale leads 32.1 to 22.3.
- DeepSeek-V3.2-Speciale is cheaper at $0.58 / $1.68 per million input/output tokens, against $2.50 / $12.50 for Nova Premier 1.0.
- Nova Premier 1.0 accepts more context: 1M tokens versus 128K.
- DeepSeek-V3.2-Speciale has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V3.2-Speciale | Nova Premier 1.0 | |
|---|---|---|
| Provider | DeepSeek | Amazon |
| Noometry Index | 39.1 | 37.7 |
| Released | 2025-12-01 | 2025-04-30 |
| Weights | Open | Proprietary |
| Context window | 128K | 1M |
| Max output | 128K | 10K |
| Input $ / M tokens | $0.58 | $2.50 |
| Output $ / M tokens | $1.68 | $12.50 |
| Results tracked | 3 | 6 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
DeepSeek-V3.2-Speciale: 39.9 (#136), Nova Premier 1.0: —
| Benchmark | DeepSeek-V3.2-Speciale | Nova Premier 1.0 |
|---|---|---|
| WeirdML | 46.7% | — |
Reasoning DeepSeek-V3.2-Speciale leads
DeepSeek-V3.2-Speciale: 32.1 (#65), Nova Premier 1.0: 22.3 (#182)
| Benchmark | DeepSeek-V3.2-Speciale | Nova Premier 1.0 |
|---|---|---|
| SimpleBench | 52.6% | — |
| Kagi LLM Benchmark | — | 44.8% |
Math Not comparable
DeepSeek-V3.2-Speciale: —, Nova Premier 1.0: 31.5 (#194)
| Benchmark | DeepSeek-V3.2-Speciale | Nova Premier 1.0 |
|---|---|---|
| Omni-MATH | — | 35% |
Knowledge Not comparable
DeepSeek-V3.2-Speciale: —, Nova Premier 1.0: 35.4 (#176)
| Benchmark | DeepSeek-V3.2-Speciale | Nova Premier 1.0 |
|---|---|---|
| MMLU-Pro | — | 72.6% |
| GPQA (HELM) | — | 51.8% |
Instruction Following Not comparable
DeepSeek-V3.2-Speciale: —, Nova Premier 1.0: 66.9 (#201)
| Benchmark | DeepSeek-V3.2-Speciale | Nova Premier 1.0 |
|---|---|---|
| IFEval | — | 80.3% |
Writing & Preference Nova Premier 1.0 leads
DeepSeek-V3.2-Speciale: 46.1 (#217), Nova Premier 1.0: 51.0 (#174)
| Benchmark | DeepSeek-V3.2-Speciale | Nova Premier 1.0 |
|---|---|---|
| EQ-Bench Creative Writing | 1276 | — |
| WildBench | — | 78.8% |
Frequently asked questions
Is DeepSeek-V3.2-Speciale better than Nova Premier 1.0?
DeepSeek-V3.2-Speciale is the stronger model overall, scoring 39.1 to 37.7 on the Noometry Index.
Which is cheaper, DeepSeek-V3.2-Speciale or Nova Premier 1.0?
DeepSeek-V3.2-Speciale is cheaper. It lists at $0.58 per million input tokens and $1.68 per million output tokens; Nova Premier 1.0 lists at $2.50 and $12.50.
Which has the bigger context window?
Nova Premier 1.0 does, with 1M tokens against 128K.
How many benchmarks do DeepSeek-V3.2-Speciale and Nova Premier 1.0 share?
0 benchmarks have published results for both models. DeepSeek-V3.2-Speciale has 3 scored results on Noometry and Nova Premier 1.0 has 6.