Everything about Kimi K2.5

What you need to know

Moonshot AI released Kimi K2.5 on January 27, 2026. It has a reasoning mode: it can "think" before answering, which helps with multi-step problems but uses more tokens, and therefore more money. It is an open-weight model: you can host it yourself on infrastructure you control, and 43 API providers listed on models.dev also offer it. That makes it 9 months old as of October 22, 2026.

We file it under open-weight models (68 models in this category on the site).

Pricing and budget

Kimi K2.5 costs $0.58 per million input tokens and $3 per million output tokens (median of 43 providers, checked on October 22, 2026).

In practice, an agent task that reads 20,000 tokens and writes 2,000 costs about $0.018, or $17.6 for 1,000 such tasks.

That is 3 times the median output price of open-weight models ($1).

To estimate your own budget, use the agent cost calculator.

Measured performance

According to results published by Epoch AI (checked on October 22, 2026), Kimi K2.5 has a capability index of 78.1 out of 100, computed on 4 benchmarks from our basket.

It is not among the top 30 of the weekly ranking.

• GPQA Diamond (PhD-level science questions): 87.6% — the best tracked model, Gemini 4 Argon, reaches 96.4%
• OTIS Mock AIME (competition mathematics problems): 92.2% — the best tracked model, GPT-5.5 Pro, reaches 100%
• SWE-bench Verified (fixing real software bugs): 73.8% — the best tracked model, Claude Opus 4.7, reaches 83.5%
• Terminal-Bench (agent tasks in a terminal): 43.2% — the best tracked model, GPT-5.5, reaches 84.7%

What it can do

Its context window reaches 262,144 tokens, roughly 393 pages of text at 500 words per page: enough for a long report or several contracts at once. A single answer can run to 98,304 tokens.

It accepts as input: text, image and video. Its knowledge stops at January 2025: beyond that, it needs web search or your documents.

It can call tools (functions, APIs) and return structured JSON output, which makes it a possible foundation for an AI agent.

The weights are published on Hugging Face (moonshotai/Kimi-K2.5). See also our list of open-source models.

What is it good for?

Given its category, Kimi K2.5 is mainly aimed at projects where you want control over your data or hosting, or want to adapt the model to your field.

Before adopting it, test it on your own examples: benchmarks measure standard tasks, not your job.

Alternatives

In the same category, the models closest to Kimi K2.5 are:

• GLM-5.3 (Zhipu AI (Z.ai), index 78.8) — see the comparison
• GLM-5 (Zhipu AI (Z.ai), index 77.2) — see the comparison
• Kimi K3 (Moonshot AI, index 80.7) — see the comparison
• Muse Spark Open (Meta, index 75.2) — see the comparison