The AI arena is free today

Open Superagent

DeepSeek-V3.2 vs DeepSeek-V4-Flash-0731

DeepSeek-V4-Flash-0731 leads the LLM Stats Score 46.1 to 33.5. DeepSeek-V4-Flash-0731 is 2.6x cheaper per token.

DeepSeek · DeepSeek · Updated for 2026

Which is better?

DeepSeek-V4-Flash-0731 leads the overall LLM Stats Score 46.1 to 33.5, ranking #23 overall.

In the 1 individual benchmarks reported for both models, DeepSeek-V4-Flash-0731 wins 1; this is a narrower head-to-head signal than the composite indexes.

On price, DeepSeek-V4-Flash-0731 is roughly 2.6x cheaper per token on a blended 3:1 input/output basis, which adds up quickly at production volume.

DeepSeek-V4-Flash-0731 also accepts a larger context window (1,048,576 input tokens), making it the stronger choice for long documents and large codebases.

Based on current LLM Stats indexes, shared benchmarks, pricing, and model metadata for 2026.

Choose DeepSeek-V3.2

  • you want predictable pricing at $0.26/M input and $0.38/M output

Choose DeepSeek-V4-Flash-0731

  • overall performance matters — it scores 46.1 and ranks #23 on LLM Stats
  • your work emphasizes reasoning and coding — it leads those capability indexes
  • you value its reported benchmark strengths — it wins 1 of 1 exact shared results
  • cost matters — it's about 2.6x cheaper per token
  • you process long inputs — it offers a 1,048,576 token context window
  • you want the most recent training data — it shipped Jul 2026

At a glance

The differences that matter most.

Core performance indexes
33.5
#86
46.1
#23
33.6
#83
43.3
#35
23.3
#67
36.2
#24
12.7
#90
33.2
#21
Cost, coverage & limits
Benchmark wins
0 of 1
1 of 1
Input price
$0.26 / M
$0.09 / M
Output price
$0.38 / M
$0.18 / M
Context window
163,840
1,048,576

Capability indexes

Additional strengths measured across groups of related public benchmarks

1 shared
Index
DeepSeek-V3.2
DeepSeek-V4-Flash-0731
10.6#114
27.1#21
Conservative TrueSkill rating · higher is betterHow scores work

Individual benchmarks

17 reported for DeepSeek-V3.2 · 9 for DeepSeek-V4-Flash-0731

1 shared

DeepSeek-V3.2 outperforms in 0 benchmarks, while DeepSeek-V4-Flash-0731 is better at 1 benchmark (Toolathlon).

DeepSeek-V4-Flash-0731 significantly outperforms across most benchmarks.

Sun Aug 30 2026 • llm-stats.com

Human preference

Blind head-to-head votes and playground preference scores

Pricing Analysis

Price comparison per million tokens

DeepSeek-V4-Flash-0731 costs less

For input processing, DeepSeek-V3.2 ($0.26/1M tokens) is 2.9x more expensive than DeepSeek-V4-Flash-0731 ($0.09/1M tokens).

For output processing, DeepSeek-V3.2 ($0.38/1M tokens) is 2.1x more expensive than DeepSeek-V4-Flash-0731 ($0.18/1M tokens).

In conclusion, DeepSeek-V3.2 is more expensive than DeepSeek-V4-Flash-0731.*

* Using a 3:1 ratio of input to output tokens

Lowest available price from all providers
Sun Aug 30 2026 • llm-stats.com
DeepSeek
DeepSeek-V3.2
Input tokens$0.26
Output tokens$0.38
Best providerDeepinfra
DeepSeek
DeepSeek-V4-Flash-0731
Input tokens$0.09
Output tokens$0.18
Best providerDeepinfra
Notice missing or incorrect data?Start an Issue

Model Size

Parameter count comparison

381.0B diff

DeepSeek-V3.2 has 381.0B more parameters than DeepSeek-V4-Flash-0731, making it 125.3% larger.

DeepSeek
DeepSeek-V3.2
685.0Bparameters
DeepSeek
DeepSeek-V4-Flash-0731
304.0Bparameters
685.0B
DeepSeek-V3.2
304.0B
DeepSeek-V4-Flash-0731

Context Window

Maximum input and output token capacity

DeepSeek-V4-Flash-0731 accepts 1,048,576 input tokens compared to DeepSeek-V3.2's 163,840 tokens. DeepSeek-V4-Flash-0731 can generate longer responses up to 384,000 tokens, while DeepSeek-V3.2 is limited to 163,840 tokens.

DeepSeek
DeepSeek-V3.2
Input163,840 tokens
Output163,840 tokens
DeepSeek
DeepSeek-V4-Flash-0731
Input1,048,576 tokens
Output384,000 tokens
Sun Aug 30 2026 • llm-stats.com

License

Usage and distribution terms

Both models are licensed under MIT.

Both models share the same licensing terms, providing consistent usage rights.

DeepSeek-V3.2

MIT

Open weights

DeepSeek-V4-Flash-0731

MIT

Open weights

Release Timeline

When each model was launched

DeepSeek-V3.2 was released on 2025-12-01, while DeepSeek-V4-Flash-0731 was released on 2026-07-31.

DeepSeek-V4-Flash-0731 is 8 months newer than DeepSeek-V3.2.

DeepSeek-V3.2

Dec 1, 2025

9 months ago

DeepSeek-V4-Flash-0731

Jul 31, 2026

1 months ago

8mo newer

Knowledge Cutoff

When training data ends

Neither model specifies a knowledge cutoff date.

Unable to compare the recency of their training data.

No cutoff dates available

Provider Availability

DeepSeek-V3.2 is available from DeepInfra, Novita, Fireworks. DeepSeek-V4-Flash-0731 is available from DeepInfra, Novita, Fireworks.

DeepSeek-V3.2

deepinfra logo
Deepinfra
Input Price:Input: $0.26/1MOutput Price:Output: $0.38/1M
novita logo
Novita
Input Price:Input: $0.27/1MOutput Price:Output: $0.40/1M
fireworks logo
Fireworks
Input Price:Input: $0.56/1MOutput Price:Output: $1.68/1M

DeepSeek-V4-Flash-0731

deepinfra logo
Deepinfra
Input Price:Input: $0.09/1MOutput Price:Output: $0.18/1M
novita logo
Novita
Input Price:Input: $0.14/1MOutput Price:Output: $0.28/1M
fireworks logo
Fireworks
Input Price:Input: $0.44/1MOutput Price:Output: $1.32/1M
* Prices shown are per million tokens

Outputs Comparison

Notice missing or incorrect data?Start an Issue discussion

Judge for yourself.

Run your own prompts against DeepSeek-V3.2 and DeepSeek-V4-Flash-0731 side-by-side, then vote on the output you prefer.

DeepSeek-V3.2
✓ Preferred
DeepSeek-V4-Flash-0731
Open in Playground

FAQ

Common questions about DeepSeek-V3.2 vs DeepSeek-V4-Flash-0731.

Which is better, DeepSeek-V3.2 or DeepSeek-V4-Flash-0731?

DeepSeek-V4-Flash-0731 leads the LLM Stats Score 46.1 to 33.5. DeepSeek-V3.2 is made by DeepSeek and DeepSeek-V4-Flash-0731 is made by DeepSeek. The best choice depends on your use case — compare their capability indexes, individual benchmarks, pricing, and limits above.

How does DeepSeek-V3.2 compare to DeepSeek-V4-Flash-0731 in benchmarks?

DeepSeek-V3.2 scores AIME 2025: 93.1%, HMMT 2025: 90.2%, MMLU-Pro: 85.0%, LiveCodeBench: 83.3%, GPQA: 82.4%. DeepSeek-V4-Flash-0731 scores Terminal-Bench 2.1: 82.7%, CyberGym: 76.7%, Toolathlon: 70.3%, DSBench-FullStack: 68.7%, DSBench-Hard: 59.6%.

Is DeepSeek-V3.2 cheaper than DeepSeek-V4-Flash-0731?

DeepSeek-V4-Flash-0731 is 2.9x cheaper for input tokens. DeepSeek-V3.2 costs $0.26/M input and $0.38/M output via deepinfra. DeepSeek-V4-Flash-0731 costs $0.09/M input and $0.18/M output via deepinfra.

What are the context window sizes for DeepSeek-V3.2 and DeepSeek-V4-Flash-0731?

DeepSeek-V3.2 supports 164K tokens and DeepSeek-V4-Flash-0731 supports 1.0M tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.

What are the main differences between DeepSeek-V3.2 and DeepSeek-V4-Flash-0731?

Key differences include LLM Stats Score (33.5 vs 46.1), context window (164K vs 1.0M), input pricing ($0.26 vs $0.09/M). See the full comparison above for benchmark-by-benchmark results.