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 Results for other workloads are historical and no longer updated.
Alibaba has launched a newer model, Qwen3 235B (Non-reasoning). We suggest considering it instead.
For more information, see comparison of Qwen3 235B (Non-reasoning) to other models and API provider benchmarks for Qwen3 235B (Non-reasoning).
Qwen2.5 Instruct 72B API Provider Benchmarking & Analysis
This analysis is intended to support you in choosing the best API provider of Qwen2.5 Instruct 72B for your use-case.
Fastest
SiliconFlow (FP8)
DeepInfraOutput speed
Total 2 providers
Lowest Latency
DeepInfraSiliconFlow (FP8)Time to first token
Total 2 providers
Lowest Price
DeepInfraSiliconFlow (FP8)Blended price (per 1M tokens)
Total 2 providers
Qwen2.5 72B is available through 2 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 SiliconFlow (FP8) (32.2 t/s) and DeepInfra (27.2 t/s).
- For latency, DeepInfra (2.35s) and SiliconFlow (FP8) (4.30s) offer the lowest time to first token.
- For pricing, DeepInfra (0.20) and SiliconFlow (FP8) (0.59) offer the lowest blended prices per 1M tokens.
- DeepInfra offers both the lowest latency and best pricing, making it attractive for cost-conscious applications. SiliconFlow (FP8) is the fastest option for throughput-intensive workloads.
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: Qwen2.5 Instruct 72B
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
Frequently Asked Questions
Common questions about Qwen2.5 Instruct 72B providers
Qwen2.5 Instruct 72B is available through 2 API providers: SiliconFlow (FP8) and DeepInfra. Each provider offers different performance characteristics and pricing.
Qwen2.5 Instruct 72B is currently available through 2 API providers that we benchmark and track.
The fastest providers for Qwen2.5 Instruct 72B by output speed are SiliconFlow (FP8) (32.2 t/s) and DeepInfra (27.2 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 Qwen2.5 Instruct 72B are DeepInfra (2.35s) and SiliconFlow (FP8) (4.30s). Lower latency means faster initial response time.
The most affordable providers for Qwen2.5 Instruct 72B by blended price are DeepInfra ($0.20 per 1M tokens) and SiliconFlow (FP8) ($0.59 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 Qwen2.5 Instruct 72B are DeepInfra ($0.36 per 1M input tokens) and SiliconFlow (FP8) ($0.59 per 1M input tokens).
The providers with the lowest output token pricing for Qwen2.5 Instruct 72B are DeepInfra ($0.40 per 1M output tokens) and SiliconFlow (FP8) ($0.59 per 1M output tokens).
Prices for Qwen2.5 Instruct 72B vary up to 3.0x across providers. The most affordable is DeepInfra at $0.20 per 1M tokens, while SiliconFlow (FP8) charges $0.59 per 1M tokens.
All 2 providers of Qwen2.5 Instruct 72B support JSON mode for structured output.
All 2 providers of Qwen2.5 Instruct 72B support function calling (tool use).
The best provider for Qwen2.5 Instruct 72B depends on your priorities: SiliconFlow (FP8) offers the highest output speed, DeepInfra has the lowest latency, and DeepInfra provides the most competitive pricing.
When choosing a provider for Qwen2.5 Instruct 72B, 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 Qwen2.5 Instruct 72B's intelligence, capabilities, modalities, and how it compares to other models, see the model overview page. View model overview