K2-V2 (medium) logo
MBZUAI Institute of Foundation Models
•

Open weights model

•

Released December 2025

K2-V2 (medium) Benchmarking und Analyse von API-Anbietern

Modellvergleich

Analyse der API-Anbieter für K2-V2 (medium) anhand von Leistungsmetriken wie Latenz (Zeit bis zum ersten Token), Ausgabegeschwindigkeit (Ausgabetokens pro Sekunde) und Preis. Zu den bewerteten API-Anbietern gehören .

Am schnellsten

Ausgabegeschwindigkeit

Insgesamt 0 Anbieter

Niedrigste Latenz

Zeit bis zum ersten Antworttoken

Insgesamt 0 Anbieter

Niedrigster Preis

Mischpreis (pro 1 Mio. Tokens)

Insgesamt 0 Anbieter

Für K2-V2 (medium) sind derzeit keine API-Anbieter verfügbar.

Für dieses Modell sind keine Anbieter-Benchmarks verfügbar.
Einzelheiten zum Modell und zum Vergleich seiner Intelligenz mit anderen Modellen finden Sie auf der Modellseite für K2-V2 (medium).

Wichtigste Ergebnisse

Geschwindigkeit

Output tokens per second · Higher is better
No data available

Ende-zu-Ende-Antwortzeit

Seconds · Lower is better
No data available

Preis

USD per 1M tokens (blended) · Lower is better
No data available

Aktualisierung: Der Standard-Workload für das Performance-Benchmarking wurde auf 10k Eingabetokens umgestellt, um Produktionsanwendungsfälle besser abzubilden. Oben können weiterhin andere Workloads ausgewählt werden.

Preise

Pricing: Cache Hit, Input, and Output

Price (USD per M Tokens) · Lower is better · 10,000 input tokens
No data available

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.

Pricing: Blended Price

Blended at 7:2:1 (cache-input-output) · USD per 1M tokens (blended) · Lower is better
No data available

Price per token, shown in USD per million tokens. Price is a blend of cache hit, input, and output token prices using the selected ratio (default 7:2:1 cache-input-output).

Output Speed vs. Price

Blended at 7:2:1 (cache-input-output) · Output speed: output tokens per second · USD per 1M tokens (blended)
Most attractive quadrant
No data available

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

Geschwindigkeit

Gemessen anhand der Ausgabegeschwindigkeit (Tokens pro Sekunde)

Output Speed: K2-V2 (medium)

Output speed: output tokens per second · 10,000 input tokens
No data available

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

Figures represent performance of the model's first-party API (e.g. OpenAI for o1) or the median across providers where a first-party API is not available (e.g. Meta's Llama models).

Latency vs. Output Speed

Latency: seconds to first token received · Output speed: output tokens per second · 10,000 input tokens
Most attractive quadrant
No data available

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

Latenz

Gemessen anhand der Zeit (in Sekunden) bis zum ersten Token

Zeit bis zum ersten Antworttoken

Seconds to first token received · Lower is better · 10,000 input tokens
No data available

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.

Figures represent median (P50) measurement over the past 72 hours to reflect sustained changes in performance.

Ende-zu-Ende-Antwortzeit

Seconds to output 500 tokens, calculated based on time to first token, 'thinking' time for reasoning models, and output speed

Ende-zu-Ende-Antwortzeit

Seconds to output 500 tokens, including reasoning model 'thinking' time · Lower is better · 10,000 input tokens
No data available

Seconds to receive a 500 token response. Key components:

  • Input time: Time to receive the first response token
  • Thinking time (only for reasoning models): Time reasoning models spend outputting tokens to reason prior to providing an answer. Amount of tokens based on the average reasoning tokens across a diverse set of 60 prompts (methodology details).
  • Answer time: Time to generate 500 output tokens, based on output speed

Wichtige Vergleichsmetriken und API-Funktionen

No results.

Häufig gestellte Fragen

Häufige Fragen zu den Anbietern von K2-V2 (medium)

K2-V2 (medium) ist derzeit über keinen der von uns bewerteten API-Anbieter verfügbar. Als Modell mit offenen Gewichten kann es selbst gehostet werden.