K2.5 supersedes K2 Thinking for most uses — it adds vision and video, Agent Swarm, and both instant and thinking modes. K2 Thinking survives on economics: its INT4-native weights are cheaper to serve for pure-text reasoning, and existing deployments have little urgency to migrate. New projects: K2.5.
Kimi K2.5 vs K2 Thinking at a glance
| Kimi K2.5 | K2 Thinking | |
|---|---|---|
| Vendor | Moonshot AI (China) | Moonshot AI (China) |
| Open weights | Yes — downloadable | Yes — downloadable |
| License | Modified MIT — free commercial use; attribution required above 100M monthly active users or $20M monthly revenue | Modified MIT — free commercial use; attribution required above 100M monthly active users or $20M monthly revenue |
| Parameters | 1T (32B active) | 1T (32B active) |
| Context window | 256K tokens | 256K tokens |
| Modalities | text, image, video | text |
| Pricing | $0.60 in / $2.50 out per 1M tokens | Varies by hosting provider (open weights) |
| Released | 2026-01-27 | 2025-11-06 |
Specs and pricing verified July 2026.
About Kimi K2.5
Kimi K2.5 is Moonshot AI's trillion-parameter multimodal agent model, released January 27, 2026. Built by continual-pretraining on ~15 trillion mixed visual and text tokens on top of the Kimi-K2 base, it natively understands text, images, and video, activates 32B parameters per token, and ships both instant and thinking modes.
Its defining feature is agentic: K2.5 introduced Agent Swarm, coordinating up to 100 specialized agents on a single task, and posted 76.8% on SWE-bench Verified — frontier-class coding from an open-weight release. The Modified MIT license is effectively free for everyone below 100M monthly users or $20M monthly revenue, at which point attribution (not payment) kicks in.
Full specs, benchmarks, and hardware guidance: the Kimi K2.5 page.
About K2 Thinking
Kimi K2 Thinking was Moonshot's November 2025 reasoning release: a trillion-parameter MoE trained for long chains of thought and sustained tool use, reportedly executing 200–300 sequential tool calls without human intervention. At release it beat proprietary frontiers on several agentic benchmarks — a genuine 'open-weight catches up' moment.
It ships INT4-native thanks to quantization-aware training, roughly halving the memory footprint typical of its class. K2.5 has since superseded it as Moonshot's flagship, but K2 Thinking remains the text-only reasoning specialist in the K2 line and a common choice on hosted inference platforms.
Full specs, benchmarks, and hardware guidance: the K2 Thinking page.
Choosing between them
Choose Kimi K2.5 for:
- Multimodal agents that need vision and video, not just text
- Frontier coding (76.8% SWE-bench Verified) on open weights
- Multi-agent orchestration workloads (Agent Swarm)
Choose K2 Thinking for:
- Long-horizon agentic tasks and heavy tool use
- Text-only reasoning workloads where K2.5's multimodality is unneeded
- Memory-efficient serving of a trillion-parameter model (INT4-native)
Frequently asked questions
Is Kimi K2.5 better than K2 Thinking?
K2.5 supersedes K2 Thinking for most uses — it adds vision and video, Agent Swarm, and both instant and thinking modes. K2 Thinking survives on economics: its INT4-native weights are cheaper to serve for pure-text reasoning, and existing deployments have little urgency to migrate. New projects: K2.5.
Which is cheaper: Kimi K2.5 or K2 Thinking?
Kimi K2.5: $0.60 in / $2.50 out per 1M tokens. K2 Thinking: Varies by hosting provider (open weights). Both are open-weight, so self-hosting costs depend on your hardware and utilization.
Can I self-host Kimi K2.5 and K2 Thinking?
Kimi K2.5: yes — weights are published under Modified MIT. K2 Thinking: yes — weights are published under Modified MIT.