This model is deprecated. We only continue performance benchmarking for the default 10k input token workload. Results for other workloads are historical and no longer updated.
DeepSeek has launched a newer model, DeepSeek V4 Pro (Non-reasoning). We suggest considering it instead.
For more information, see comparison of DeepSeek V4 Pro (Non-reasoning) to other models and API provider benchmarks for DeepSeek V4 Pro (Non-reasoning).
DeepSeek V3.2 (Non-reasoning) API Provider Benchmarking & Analysis
This analysis is intended to support you in choosing the best API provider of DeepSeek V3.2 (Non-reasoning) for your use-case.
Fastest
Azure
Amazon
FriendliAI
GMI
DigitalOceanOutput speed
Total 11 providers
Lowest Latency
FriendliAI
DeepInfraAmazonDigitalOceanAzureTime to first token
Total 11 providers
Lowest Price
DeepInfraDigitalOcean
SiliconFlow (FP8)GMI
NovitaBlended price (per 1M tokens)
Total 11 providers
DeepSeek V3.2 (Non-reasoning) is available through 11 API providers, each offering different performance characteristics and pricing. Below is a comparison of the key metrics across providers.
- For output speed, the top providers are Azure (162.1 t/s), Amazon (76.3 t/s), and FriendliAI (58.0 t/s). Speed varies significantly across providers, with a 637% difference between the fastest and slowest.
- For latency, FriendliAI (1.27s), DeepInfra (1.35s), and Amazon (1.51s) offer the lowest time to first token.
- For pricing, DeepInfra (0.18), DigitalOcean (0.18), and SiliconFlow (FP8) (0.19) offer the lowest blended prices per 1M tokens. Prices vary up to 17.4x across providers.
Highlights
Pricing
Pricing: Cache Hit, Input, and Output
Pricing: Blended Price
Pricing: Cache Discount
Output Speed vs. Price
Speed
Measured by Output Speed (tokens per second)
Output Speed: DeepSeek V3.2 (Non-reasoning)
Latency vs. Output Speed
Latency
Measured by Time (seconds) to First Token
Time to First Token
End-to-End Response Time
Seconds to output 500 tokens, calculated based on time to first token, 'thinking' time for reasoning models, and output speed
End-to-End Response Time
Key Comparison Metrics & API Features
| GMI | 164k | Open | -- | 50 | 2.67 | 12.63 | -- | ||
| Novita | 164k | Open | -- | 39 | 2.98 | 15.64 | -- | ||
 SambaNova | 8.19k | Open | -- | -- | -- | -- | -- | ||
 Fireworks | 164k | Open | -- | -- | -- | -- | -- | ||
| Microsoft Azure | 128k | Open | -- | 157 | 1.86 | 5.04 | -- | ||
| Amazon Bedrock | 128k | Open | -- | 66 | 1.47 | 9.05 | -- | ||
| DeepInfra | 164k | Open | -- | 24 | 1.19 | 22.36 | -- | ||
 DeepSeek | 128k | Open | -- | -- | -- | -- | -- | ||
| DigitalOcean | 128k | Open | -- | 44 | 1.48 | 12.84 | -- | ||
| FriendliAI | 164k | Open | -- | 62 | 1.14 | 9.26 | -- | ||
| SiliconFlow | 164k | Open | -- | 22 | 2.68 | 24.99 | -- | ||
Frequently Asked Questions
Common questions about DeepSeek V3.2 (Non-reasoning) providers
DeepSeek V3.2 (Non-reasoning) is available through 11 API providers: GMI, Novita, SambaNova, Fireworks, Azure, Amazon, DeepInfra, DeepSeek, DigitalOcean, FriendliAI, and SiliconFlow (FP8). Each provider offers different performance characteristics and pricing.
DeepSeek V3.2 (Non-reasoning) is currently available through 11 API providers that we benchmark and track.
The fastest providers for DeepSeek V3.2 (Non-reasoning) by output speed are Azure (162.1 t/s), Amazon (76.3 t/s), and FriendliAI (58.0 t/s). Output speed measures how quickly tokens are generated after the model starts responding.
The providers with the lowest time to first token for DeepSeek V3.2 (Non-reasoning) are FriendliAI (1.27s), DeepInfra (1.35s), and Amazon (1.51s). Lower latency means faster initial response time.
The most affordable providers for DeepSeek V3.2 (Non-reasoning) by blended price are DeepInfra ($0.18 per 1M tokens), DigitalOcean ($0.18 per 1M tokens), and SiliconFlow (FP8) ($0.19 per 1M tokens). Blended price uses a 7:2:1 cache hit/input/output token ratio.
The providers with the lowest input token pricing for DeepSeek V3.2 (Non-reasoning) are GMI ($0.21 per 1M input tokens), DigitalOcean ($0.25 per 1M input tokens), and DeepInfra ($0.26 per 1M input tokens).
Output speed for DeepSeek V3.2 (Non-reasoning) varies significantly across providers. Azure is the fastest at 162.1 t/s, which is 7.4x faster than SiliconFlow (FP8) at 22.0 t/s.
10 of 11 providers support JSON mode for DeepSeek V3.2 (Non-reasoning): GMI, Novita, SambaNova, Fireworks, Azure, DeepInfra, DeepSeek, DigitalOcean, FriendliAI, and SiliconFlow (FP8).
All 11 providers of DeepSeek V3.2 (Non-reasoning) support function calling (tool use).
The best provider for DeepSeek V3.2 (Non-reasoning) depends on your priorities: Azure offers the highest output speed, FriendliAI has the lowest latency, and DeepInfra provides the most competitive pricing.
When choosing a provider for DeepSeek V3.2 (Non-reasoning), consider: output speed (for throughput-intensive tasks), latency (for interactive applications requiring quick first responses), pricing (for cost-sensitive workloads), and API features like JSON mode or function calling.
Yes, provider performance can vary over time due to infrastructure changes, load balancing, and updates. We continuously benchmark all providers and display historical performance trends in the "Over Time" charts.
For information about DeepSeek V3.2 (Non-reasoning)'s intelligence, capabilities, modalities, and how it compares to other models, see the model overview page. View model overview