Compare · ModelsLive · 2 picked · head to head
DeepSeek V3.2 vs Gemma 4 26B A4B
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
DeepSeek V3.2 wins on 5/8 benchmarks
DeepSeek V3.2 wins 5 of 8 shared benchmarks. Leads in speed · knowledge · general.
Category leads
speed·DeepSeek V3.2arena·Gemma 4 26B A4B knowledge·DeepSeek V3.2general·DeepSeek V3.2math·DeepSeek V3.2
Hype vs Reality
Attention vs performance
DeepSeek V3.2
#142 by perf·no signal
Gemma 4 26B A4B
#151 by perf·#7 by attention
Best value
Gemma 4 26B A4B
2.3x better value than DeepSeek V3.2
DeepSeek V3.2
136.3 pts/$
$0.35/M
Gemma 4 26B A4B
314.5 pts/$
$0.15/M
Vendor risk
Mixed exposure
One or more vendors flagged
DeepSeek
$3.4B·Tier 1
Google DeepMind
$4.20T·Tier 1
Head to head
8 benchmarks · 2 models
DeepSeek V3.2Gemma 4 26B A4B
Artificial Analysis · Quality Index
DeepSeek V3.2 leads by +25.0
DeepSeek V3.2
41.7
Gemma 4 26B A4B
16.7
Chatbot Arena Elo · Coding
Gemma 4 26B A4B leads by +33.6
DeepSeek V3.2
1324.8
Gemma 4 26B A4B
1358.5
Chatbot Arena Elo · Overall
Gemma 4 26B A4B leads by +12.7
DeepSeek V3.2
1424.8
Gemma 4 26B A4B
1437.5
Chess Puzzles
DeepSeek V3.2 leads by +8.4
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
DeepSeek V3.2
9.5
Gemma 4 26B A4B
1.1
Dtbench
DeepSeek V3.2 leads by +17.8
DeepSeek V3.2
76.0
Gemma 4 26B A4B
58.2
GPQA diamond
DeepSeek V3.2 leads by +13.6
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
DeepSeek V3.2
77.9
Gemma 4 26B A4B
64.3
Lmca
Gemma 4 26B A4B leads by +1.0
DeepSeek V3.2
33.9
Gemma 4 26B A4B
34.9
OTIS Mock AIME 2024-2025
DeepSeek V3.2 leads by +5.6
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
DeepSeek V3.2
87.8
Gemma 4 26B A4B
82.2
Full benchmark table
| Benchmark | DeepSeek V3.2 | Gemma 4 26B A4B |
|---|---|---|
Artificial Analysis · Quality Index | 41.7 | 16.7 |
Chatbot Arena Elo · Coding | 1324.8 | 1358.5 |
Chatbot Arena Elo · Overall | 1424.8 | 1437.5 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 9.5 | 1.1 |
Dtbench | 76.0 | 58.2 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 77.9 | 64.3 |
Lmca | 33.9 | 34.9 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 87.8 | 82.2 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.28 | $0.42 | 164K tokens (~82 books) | $3.15 | |
| $0.07 | $0.23 | 262K tokens (~131 books) | $1.07 |