GPT-5.6 Luna vs Kimi K2.5
Both are capability tier 5.
| GPT-5.6 Luna | Kimi K2.5 | |
|---|---|---|
| Vendor | OpenAI | Moonshot |
| Capability tier | 5/8 | 5/8 |
| Size class | FAST | BALANCED |
| Parameters | — | 1,000B |
| Context window | 1,050,000 | — |
| Input $/M | $0.20 | — |
| Output $/M | $1.20 | — |
| Cached input $/M | $0.10 | — |
FAQ
Is GPT-5.6 Luna cheaper than Kimi K2.5?
Pricing isn't listed for both models — at least one is open-weights (self-host). Use the Mintok sizing report to compare $/M-token.
GPT-5.6 Luna vs Kimi K2.5: which has a bigger context window?
Both list the same context window (1,050,000 tokens).
Which is more capable, GPT-5.6 Luna or Kimi K2.5?
Both sit at the same capability tier (5/8) — pick on price, context, or benchmark them on your prompts.
Settle it on your own prompts
List prices only go so far — benchmark GPT-5.6 Luna and Kimi K2.5 on your actual workload for measured cost, latency and quality.
Pricing and specs maintained by Mintok's catalog watchers from public sources. Not investment advice. Methodology.