Models

DeepSeek V3.2 vs DeepSeek V4 Flash

Compare DeepSeek V3.2 from DeepSeek and DeepSeek V4 Flash from DeepSeek on key metrics including benchmarks, price, context length, and other model features. Access both models and hundreds of others through the AIHubMix API.

DeepSeekDeepSeek V3.2DeepSeekDeepSeek V4 Flash
DeepSeek logo
DeepSeek V3.2
DeepSeek · text → text

DeepSeek V3.2 is DeepSeek's text large language model, designed for reasoning tasks and agentic tool use. It harmonizes high computational efficiency with strong reasoning performance through DeepSeek Sparse Attention (DSA), while providing native support for tools, function calling, and structured outputs.

Input$0.274 /M
Output$0.411 /M
Cache read$0.0274 /M
DeepSeek logo
DeepSeek V4 Flash
DeepSeek · text → text

DeepSeek V4 Flash 0423 is DeepSeek's efficiency-optimized Mixture-of-Experts text model, designed for fast inference and workflows requiring tool use, function calling, and structured outputs. It features 284B total parameters with 13B activated parameters and supports a 1M-token context window with cost-effective pricing.

Input$0.464 /M
Output$0.928 /M
Cache read$0.0039 /M

Pricing & Specifications

Prices are per million tokens. Time to First Token and throughput are rolling averages measured on AIHubMix.

DeepSeek V3.2
DeepSeek V4 Flash
Input /M
$0.274
$0.464
Output /M
$0.411
$0.928
Cache read /M
$0.0274
$0.0039
Context length
163,840
1,000,000
Max output
0
0
Time to First Token
0.7 s
0.7 s
Throughput
59.0 tok/s
59.0 tok/s
Modalities
text
text
Supported Parameters
toolsfunction callingstructured outputs
toolsfunction callingstructured outputs
API Formats
Released
-
-

Promotional prices show the discounted rate; see each model page for promotion windows.

Activity Past 30 Days

Daily traffic served through AIHubMix — how demand for each model is trending.

deepseek-v3.2deepseek-v4-flash

Tokens / day

-

Requests / day

-

Performance Past 3 Days

Measured on real AIHubMix traffic, hourly buckets. Gaps mean no traffic in that hour.

deepseek-v3.2deepseek-v4-flash

Throughput (tok/s)

-

TTFT (s)

-

Uptime (%)

-

LMArena Benchmarks

LMArena ratings by capability (Bradley-Terry, commonly called Elo). Higher is better.

Text
deepseek-v3.2deepseek-v4-flash
1380142014601500
Overall
14241431
Coding
14491452
Math
14281437
Instruction following
14131420
Longer query
14291435
Chinese
14601467
English
14351439

Source: LMArena (arena.ai) leaderboard, imported by AIHubMix. Models without published ratings are omitted per chart.

Cost calculator

Estimate your monthly bill for the same workload on each model.

DeepSeek V3.2
$22.61 /mo
DeepSeek V4 Flash
$41.76 /mo

Monthly = daily × 30. Discounted rates applied where a promotion is active.

FAQ

Which is cheaper: DeepSeek V3.2, DeepSeek V4 Flash?

DeepSeek V3.2: $0.411/M output tokens; DeepSeek V4 Flash: $0.928/M. Use the cost calculator above to estimate your own workload.

How do their coding arena scores compare?

DeepSeek V4 Flash: 1452; DeepSeek V3.2: 1449 (LMArena coding leaderboard).

Which responds faster?

DeepSeek V3.2: 0.7s time to first token measured on AIHubMix; see the live performance charts above for how each model behaves across the day.

How large is each context window?

DeepSeek V3.2 accepts 163,840 and DeepSeek V4 Flash accepts 1,000,000 input tokens.

Which one generates tokens faster?

DeepSeek V3.2 at 59.0 tok/s and DeepSeek V4 Flash at 59.0 tok/s, measured as output throughput on AIHubMix — a separate metric from time to first token.

What inputs and capabilities does each model support?

DeepSeek V3.2 accepts text input and supports tool calling, function calling and structured outputs; DeepSeek V4 Flash accepts text input and supports tool calling, function calling and structured outputs.

Can I call DeepSeek V3.2 and DeepSeek V4 Flash with the same API key?

Yes. AIHubMix serves every model on this page behind one OpenAI-compatible endpoint, so switching between them is a one-line change to the model field — no second account, key or SDK.

Popular comparisons

Related model match-ups readers also look at.