Kimi K3 vs MAI-Code-1.1-Flash: Specs & Benchmark Comparison
Kimi K3 is developed by Moonshot AI, while MAI-Code-1.1-Flash comes from Microsoft. Kimi K3 was released in July 2026, and MAI-Code-1.1-Flash followed a month later in August 2026. Kimi K3 is the larger model at roughly 2.8 trillion parameters, against 138 billion for MAI-Code-1.1-Flash.
We have 31 benchmark results for Kimi K3 and 2 benchmark results for MAI-Code-1.1-Flash, but they overlap on a single test: Terminal-Bench 2.1, where Kimi K3 scores 88.3% against 62.9%.
MAI-Code-1.1-Flash is the cheaper API at $0.20 per million input tokens and $1.20 per million output tokens, roughly 15 times cheaper than Kimi K3 at $3 and $15. Kimi K3 takes the larger context window at 1M tokens, compared with 256K for MAI-Code-1.1-Flash. Both accept text and images as input. On tooling, only Kimi K3 supports function calling and only Kimi K3 offers structured output.
| Characteristic | Kimi K3 | MAI-Code-1.1-Flash |
|---|---|---|
| Company | Moonshot AI | Microsoft |
| Release Date | July 16, 2026 | August 11, 2026 |
| Parameters | 2.8T | 138B |
| Multimodal | Yes | Yes |
| Context (input) | 1.0M | 256K |
| Context (output) | 1.0M | — |
| Input Price / 1M | $3.00 | $0.20 |
| Output Price / 1M | $15.00 | $1.20 |
| Average Score | 67.8% | 67.8% |
| Benchmarks | ||
| Terminal-Bench 2.1 | 88.3% | 62.9% |
Visual Benchmark Comparison
Verdict
MAI-Code-1.1-Flash leads in 2 out of 4 comparison categories.
Both models show comparable average scores: Kimi K3 — 0.7, MAI-Code-1.1-Flash — 0.7.
MAI-Code-1.1-Flash is 12.9x cheaper: input $0.20/1M vs $3.00/1M tokens.
Kimi K3 supports a larger context: 1M vs 256K tokens.
MAI-Code-1.1-Flash is newer: released 8/11/2026 vs 7/16/2026.
More About These Models
Related Comparisons
Frequently Asked Questions
Which is better for coding — Kimi K3 or MAI-Code-1.1-Flash?
Which model is cheaper — Kimi K3 or MAI-Code-1.1-Flash?
Which has a larger context window — Kimi K3 or MAI-Code-1.1-Flash?
The Kimi K3 and MAI-Code-1.1-Flash comparison is updated for 2026. Data includes benchmark results, API pricing, context window size and other specifications. For more detailed information, visit the Kimi K3 or MAI-Code-1.1-Flash page. See also the complete list of AI model comparisons.