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 Results for other workloads are historical and no longer updated.
Kimi has launched a newer release, Kimi K2.6. We suggest considering it instead.
Kimi K2.5 (Reasoning) API Provider Benchmarking & Analysis
This analysis is intended to support you in choosing the best API provider of Kimi K2.5 (Reasoning) for your use-case.
Fastest
Azure
Novita
DeepInfra
Amazon
SiliconFlow (FP8)Output speed
Total 5 providers
Lowest Latency
AzureNovitaDeepInfraAmazonSiliconFlow (FP8)Time to first answer token
Total 5 providers
Lowest Price
DeepInfraSiliconFlow (FP8)NovitaAzureAmazonBlended price (per 1M tokens)
Total 5 providers
Kimi K2.5 is available through 5 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 (166.8 t/s), Novita (53.4 t/s), and DeepInfra (40.6 t/s). Speed varies significantly across providers, with a 495% difference between the fastest and slowest.
- For latency, Azure (19.48s), Novita (57.46s), and DeepInfra (74.59s) offer the lowest time to first answer token.
- For pricing, DeepInfra (0.36), SiliconFlow (FP8) (0.36), and Novita (0.49) offer the lowest blended prices per 1M tokens. Prices vary up to 2.3x across providers.
- Azure offers the best performance with both the highest speed and lowest latency. For cost optimization, DeepInfra provides the most competitive pricing.
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: Kimi K2.5 (Reasoning)
Latency vs. Output Speed
Latency
Measured by Time (seconds) to First Token
Time to First Answer 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 Kimi K2.5 (Reasoning) providers
Kimi K2.5 (Reasoning) is available through 5 API providers: DeepInfra, SiliconFlow (FP8), Azure, Amazon, and Novita. Each provider offers different performance characteristics and pricing.
Kimi K2.5 (Reasoning) is currently available through 5 API providers that we benchmark and track.
The most affordable providers for Kimi K2.5 (Reasoning) by blended price are DeepInfra ($0.36 per 1M tokens), SiliconFlow (FP8) ($0.36 per 1M tokens), and Novita ($0.49 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 Kimi K2.5 (Reasoning) are DeepInfra ($0.45 per 1M input tokens), SiliconFlow (FP8) ($0.45 per 1M input tokens), and Azure ($0.60 per 1M input tokens).
The providers with the lowest output token pricing for Kimi K2.5 (Reasoning) are DeepInfra ($2.25 per 1M output tokens), SiliconFlow (FP8) ($2.25 per 1M output tokens), and Azure ($3.00 per 1M output tokens).
Output speed for Kimi K2.5 (Reasoning) varies significantly across providers. Azure is the fastest at 166.8 t/s, which is 5.9x faster than SiliconFlow (FP8) at 28.0 t/s.
4 of 5 providers support JSON mode for Kimi K2.5 (Reasoning): DeepInfra, SiliconFlow (FP8), Azure, and Novita.
All 5 providers of Kimi K2.5 (Reasoning) support function calling (tool use).
When choosing a provider for Kimi K2.5 (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 Kimi K2.5 (Reasoning)'s intelligence, capabilities, modalities, and how it compares to other models, see the model overview page. View model overview