Arena Elo 1431Human-preference rating · LMArena (shown where the model is ranked)
Capabilities
7-day heat trend
−20.2%Pricing breakdown
Typical 3:1 output-to-input mix, per 1M tokens
Estimated monthly cost by workload
Market position
- Cheaper than 27% of tracked models
- Faster than 18% of tracked models
- Efficiency rank: #246 of 1105
Best suited for
Mixed text, image, audio and document workloads that benefit from one model across modalities.
About kimi-k2.5
B.AI routed catalog row for Kimi K2.5 with multimodal agent metadata and credit-based cache pricing.
kimi-k2.5 is a Multimodal model from B.AI (US). HotON.ai tracks it at $0.23 per 1M input tokens and $3.00 per 1M output tokens, with a 256K-token context window, ~130 tokens/sec throughput and 97.7% availability. Its composite efficiency score is 89/100 at an estimated $0.002 per successful task.
Compare kimi-k2.5
Related market news
Frequently asked questions
How much does kimi-k2.5 cost per 1M tokens?+
kimi-k2.5 is tracked at $0.23 per 1M input tokens and $3.00 per 1M output tokens. A typical 3:1 output-to-input workload blends to roughly $2.31 per 1M tokens. Figures are illustrative demo data.
What is kimi-k2.5 best for?+
Mixed text, image, audio and document workloads that benefit from one model across modalities.
How fast is kimi-k2.5?+
kimi-k2.5 delivers about 130 tokens/sec with 97.7% tracked availability, suitable for latency-sensitive, real-time applications.
Is kimi-k2.5 cheaper than other AI models?+
Within the HotON.ai tracked set, kimi-k2.5 is cheaper than 27% of models on input price and ranks #246 of 1105 by overall efficiency.
Related models
Pricing is real (via the TestKey catalog, updated daily). Quality (Arena Elo) is real where the model is ranked on LMArena. Speed, availability and efficiency are modeled estimates.