Kimi K3 vs Kimi K2.7 Code
Side-by-side comparison of API pricing, specs, benchmarks, and capabilities
| Specification |
by Moonshot
|
by Moonshot
3 providers:
Fireworks $0.950/$4.00
Moonshot $0.950/$4.00
Together $0.950/$4.00
|
|---|---|---|
| Overview | ||
| Status | Current flagship Open weights | Current coding Open weights |
| Released | Jul 16, 2026 | Jun 12, 2026 |
| Pricing per million tokens | ||
| Input | $3.00/Mtok | $0.950/Mtok |
| Output | $15.00/Mtok | $4.00/Mtok |
| Blended avg | $6.00/Mtok | $1.71/Mtok |
| Cached input | $0.300/Mtok | $0.190/Mtok |
| Specifications | ||
| Context window | 1M tokens | 262K tokens |
| Parameters | Proprietary | Proprietary |
| Speed (TPS) | — | — |
| Modalities | ||
| Input |
textimage
|
textimagevideo
|
| Benchmarks sources: Vellum LLM Leaderboard / Moonshot model card | ||
| SWE-Bench Verified | 93.4 | 50 vendor |
| HumanEval | 76.8 | 89 vendor |
| Avg benchmark score | 83.3 | — |
| Perf / dollar | 13.9 | — |
| GPQA Diamond | 93.5 | — |
| Terminal-Bench 2.1 | 88.3 | — |
| MCP Atlas | 84.2 | — |
| BrowseComp | 91.2 | — |
| Humanity's Last Exam | 56 | — |
| Independent composite scores each on its own scale — not part of the average above | ||
| AA Intelligence Index | 59.7 | 43 |
| AA Agentic Index | 54.3 | 30.3 |
| AA-Omniscience Index (−100–100) | 19.7 | -10.2 |
| GDPval-AA v2 | 1668 ELO | — |
| LMSYS Chatbot Arena | 1485.3 ELO | — |
| Providers | ||
| Available from |
Moonshot — $3.00/$15.00/Mtok
|
Fireworks — $0.950/$4.00/Mtok
Moonshot — $0.950/$4.00/Mtok
Together — $0.950/$4.00/Mtok
|
Cost at scale
1M tokens · 50/50 input/output| Volume | Kimi K3 | Kimi K2.7 Code | Savings |
|---|---|---|---|
| 1M tokens | $6 | $1.71 | $4.29 (71.5%) |
| 10M tokens | $60 | $17.13 | $42.88 (71.5%) |
| 100M tokens | $600 | $171.25 | $428.75 (71.5%) |
| 1000M tokens | $6000 | $1712.5 | $4287.5 (71.5%) |
When to pick which
distilled from the pricing and spec data above- your workload re-reads context (agents, RAG, long chats): cached input costs $0.300/Mtok — 90% off its list input price
- you need the longer context: 1M tokens vs 262K (4×)
- cost dominates: $1.71/Mtok blended vs $6.00 — 71.5% less on the same 50/50 token mix
- provider flexibility: available from 3 providers, so you can price-shop hosts or fail over
The cheaper option here — Kimi K2.7 Code costs $1.71/Mtok blended on Moonshot.
Summary
Kimi K3 by Moonshot costs $3.00/Mtok input and $15.00/Mtok output, with a 1M-token context window. It supports text, image input.
Kimi K2.7 Code by Moonshot costs $0.950/Mtok input and $4.00/Mtok output, with a 262K-token context window. It supports text, image, video input and is available from 3 providers.
On a blended cost basis, Kimi K2.7 Code is 71.5% cheaper than Kimi K3.
The two aren't directly comparable on average benchmark score: Kimi K3 has published per-benchmark results, while Kimi K2.7 Code does not yet — it is measured today on independent composites (see the table above).
Note: Pricing is per million tokens. Actual costs vary with usage patterns, prompt caching, and batch discounts. Always verify against official provider pricing pages.