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 Results for other workloads are historical and no longer updated.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 K3. We suggest considering it instead.
Kimi K2.6 (Reasoning) API Provider Benchmarking & Analysis
This analysis is intended to support you in choosing the best API provider of Kimi K2.6 (Reasoning) for your use-case.
Fastest
Nebius
Azure
CoreWeave
Parasail
NovitaOutput speed
Total 11 providers
Lowest Latency
NebiusAzureCoreWeaveParasail
KimiTime to first answer token
Total 11 providers
Lowest Price
CoreWeave
SiliconFlow (FP8)
DeepInfra (FP4)NovitaParasailBlended price (per 1M tokens)
Total 11 providers
Kimi K2.6 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 Nebius (250.1 t/s), Azure (236.4 t/s), and CoreWeave (163.6 t/s). Speed varies significantly across providers, with a 808% difference between the fastest and slowest.
- For latency, Nebius (19.34s), Azure (20.27s), and CoreWeave (28.28s) offer the lowest time to first answer token.
- For pricing, CoreWeave (0.58), SiliconFlow (FP8) (0.59), and DeepInfra (FP4) (0.60) offer the lowest blended prices per 1M tokens. Prices vary up to 2.2x across providers.
- Nebius offers the best performance with both the highest speed and lowest latency. For cost optimization, CoreWeave provides the most competitive pricing.
Pricing
Pricing: Cache Hit, Input, and Output
Cost per Intelligence Index Task
Output Speed vs. Price
Speed
Measured by Output Speed (tokens per second)
Output Speed: Kimi K2.6 (Reasoning)
Time per Intelligence Index Task
Cost per Task vs. Time per Task
Latency
Measured by Time (seconds) to First Token
Time to First Answer Token
Cache Behaviour
Cache Hit Rate
Cost per Task vs. Cache Hit Rate
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
 Baseten | 262k | Open | $0.74 | -- | -- | -- | -- | ||
| CoreWeave | 262k | Open | $1.11 | 178 | 1.04 | 28.87 | 25.02 | ||
| Novita | 262k | Open | $0.67 | 74 | 3.88 | 70.39 | 59.80 | ||
 GMI | 262k | Open | -- | -- | -- | -- | -- | ||
 Fireworks | 262k | Open | $0.52 | -- | -- | -- | -- | ||
| SiliconFlow | 262k | Open | $0.61 | 28 | 1.62 | 176.35 | 157.09 | ||
| DeepInfra | 262k | Open | $0.51 | 105 | 0.67 | 47.73 | 42.31 | ||
| Parasail | 262k | Open | $0.68 | 114 | 1.66 | 45.18 | 39.12 | ||
| Nebius | 262k | Open | $1.91 | 245 | 1.57 | 21.79 | 18.18 | ||
| Microsoft Azure | 262k | Open | $1.91 | 238 | 1.43 | 22.27 | 18.74 | ||
| Kimi | 256k | Open | $0.80 | 77 | 2.71 | 66.73 | 57.56 | ||
Frequently Asked Questions
Common questions about Kimi K2.6 (Reasoning) providers
Kimi K2.6 (Reasoning) is currently available through 11 API providers that we benchmark and track.
The most affordable providers for Kimi K2.6 (Reasoning) by blended price are CoreWeave ($0.58 per 1M tokens), SiliconFlow (FP8) ($0.59 per 1M tokens), and DeepInfra (FP4) ($0.60 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 K2.6 (Reasoning) are CoreWeave ($0.65 per 1M input tokens), DeepInfra (FP4) ($0.75 per 1M input tokens), and Parasail ($0.75 per 1M input tokens).
The providers with the lowest output token pricing for Kimi K2.6 (Reasoning) are Novita ($3.40 per 1M output tokens), SiliconFlow (FP8) ($3.40 per 1M output tokens), and CoreWeave ($3.41 per 1M output tokens).
Output speed for Kimi K2.6 (Reasoning) varies significantly across providers. Nebius is the fastest at 250.1 t/s, which is 9.1x faster than SiliconFlow (FP8) at 27.5 t/s.
All 11 providers of Kimi K2.6 (Reasoning) support JSON mode for structured output.
All 11 providers of Kimi K2.6 (Reasoning) support function calling (tool use).
When choosing a provider for Kimi K2.6 (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 K2.6 (Reasoning)'s intelligence, capabilities, modalities, and how it compares to other models, see the model overview page. View model overview