This model is deprecated. We still benchmark its performance, but less often: every 18 hours instead of every 6.

Kimi has launched a newer release, Kimi K2 0905. We suggest considering it instead.

Kimi

•

Open weights model

•

Released July 2025

Kimi K2 API Provider Benchmarking & Analysis

Model Comparison

This analysis is intended to support you in choosing the best API provider of Kimi K2 for your use-case.

Lowest Time per Task

Time per Intelligence Index task (minutes)

Total 1 provider

Lowest Cost per Task

Cost per Intelligence Index task

Total 1 provider

Kimi K2 is available through Novita.

Time per Task

Inference time (minutes) per Intelligence Index task · Lower is better

No data available

Cost per Task

Weighted average cost (USD) per Intelligence Index task · Lower is better

No data available

Pricing

Cost per Intelligence Index Task: Kimi K2

Weighted average cost (USD) per Intelligence Index task · Segmented by token type · Lower is better

No data available

Weighted average cost per Intelligence Index task. Each evaluation’s cost is calculated from input, cache hit, cache write, reasoning, and answer token prices, divided by task count, and weighted by its Intelligence Index weight.

Pricing: Cache Hit, Input, and Output: Kimi K2

Price (USD per M Tokens) · Lower is better

Price per token for cached prompts (previously processed), typically offering a significant discount compared to regular input price, represented as USD per million tokens. The values shown here are the cache hit price; cache write and cache storage are billed separately and vary by provider — see "Cache pricing by provider" for detail.

Output Speed vs. Cost per Task: Kimi K2

Output speed: output tokens divided by total request time, including time to first token · Weighted average cost (USD) per Intelligence Index task

Most attractive quadrant

No data available

Output tokens, including reasoning tokens, received per second over a request's full duration, from sending the request to receiving the final token. The headline value averages each Performance v3 workload over the past 5 days, then averages the workloads.

Speed

Average of the single-request and multi-turn agentic workloads measured

Output Speed: Kimi K2

Output tokens divided by total request time · Includes time to first token · Average of the single-request and multi-turn agentic workloads measured · Higher is better

Speed and latency metrics average each endpoint's runs over the past 5 days, and the error rate pools them. We run the benchmark every 6 hours, and endpoints of deprecated models every 18 hours.

Latency

Average of the single-request and multi-turn agentic workloads measured

Time to First Answer Token: Kimi K2

Seconds until the first answer token after any reasoning · Average of the single-request and multi-turn agentic workloads measured · Lower is better

Time to first answer token received, in seconds, after API request sent. For reasoning models, this includes the 'thinking' time of the model before providing an answer. For models which do not support streaming, this represents time to receive the completion.

Speed and latency metrics average each endpoint's runs over the past 5 days, and the error rate pools them. We run the benchmark every 6 hours, and endpoints of deprecated models every 18 hours.

Reliability

Share of requests the endpoint failed, across all the single-request and multi-turn agentic workloads measured

Endpoint Error Rate: Kimi K2

Share of requests the endpoint failed with an error, a rate limit or a timeout · Across all the single-request and multi-turn agentic workloads measured · Lower is better

Share of requests over the past 5 days that the endpoint failed with an error, a rate limit or a timeout. Failures on our side are left out of both counts. Across several workloads, the failures and requests are pooled before dividing.

Speed and latency metrics average each endpoint's runs over the past 5 days, and the error rate pools them. We run the benchmark every 6 hours, and endpoints of deprecated models every 18 hours.

Caching

Cache Hit Rate: Kimi K2

Median across multi-turn agentic session runs in the last four weeks, updated Oct 26, 2026 · Higher is better

The share of cacheable input tokens served from the provider's prompt cache, measured during multi-turn agentic sessions. We regularly measure each endpoint and report a representative rate based on recent measurements. Higher values mean more of the input is billed at the cheaper cache-hit price, where the provider offers one, which lowers the effective cost of agentic workloads. For further details, see our methodology page.

Cost per Task vs. Cache Hit Rate: Kimi K2

Weighted average cost (USD) per Intelligence Index task · Cache hit rate: median of the last four weeks, updated Oct 26, 2026

Most attractive quadrant

No data available

Key Comparison Metrics & API Features



Novita	--	--	9.53	2.6%	131k			Open

Frequently Asked Questions

Common questions about Kimi K2 providers

Kimi K2 is available through 1 API provider: Novita. Each provider offers different performance characteristics and pricing.

Kimi K2 is currently available through 1 API provider that we benchmark and track.

The provider with the lowest time to first answer token for Kimi K2 is Novita (9.53s). Lower latency means faster initial response time.

The most affordable provider for Kimi K2 by blended price is Novita ($0.74 per 1M tokens). Blended price uses a 7:2:1 cache hit/input/output token ratio.

The provider with the lowest input token pricing for Kimi K2 is Novita ($0.57 per 1M input tokens).

The provider with the lowest output token pricing for Kimi K2 is Novita ($2.30 per 1M output tokens).

The only provider of Kimi K2 supports JSON mode for structured output.

The only provider of Kimi K2 supports function calling (tool use).

When choosing a provider for Kimi K2, consider: time per task (how quickly the provider completes an Intelligence Index task), 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's intelligence, capabilities, modalities, and how it compares to other models, see the model overview page. View model overview

Kimi K2 API Provider Benchmarking & Analysis

Lowest Time per Task

Lowest Cost per Task

Time per Task

Cost per Task

Pricing

Cost per Intelligence Index Task: Kimi K2

Cost per Intelligence Index Task

Pricing: Cache Hit, Input, and Output: Kimi K2

Cache Hit

Output Speed vs. Cost per Task: Kimi K2

End-to-End Output Speed

Speed

Output Speed: Kimi K2

End-to-End Output Speed

Time Period

Latency

Time to First Answer Token: Kimi K2

Time to First Answer Token

Time Period

Reliability

Endpoint Error Rate: Kimi K2

Endpoint Error Rate

Time Period

Caching

Cache Hit Rate: Kimi K2

What Cache Hit Rate Shows

Cost per Task vs. Cache Hit Rate: Kimi K2

Cost per Intelligence Index Task

What Cache Hit Rate Shows

Key Comparison Metrics & API Features

Frequently Asked Questions

Where can I access Kimi K2?

How many API providers offer Kimi K2?

Which provider has the lowest latency for Kimi K2?

Which provider is cheapest for Kimi K2?

Which provider has the lowest input price for Kimi K2?

Which provider has the lowest output price for Kimi K2?

Which Kimi K2 providers support JSON mode?

Which Kimi K2 providers support function calling?

How do I choose a provider for Kimi K2?

Does provider performance for Kimi K2 change over time?

What are the overall capabilities of Kimi K2?