LLM Comparison
Gemini 3.5 Flash-Lite vs Kimi K2 Thinking
Side-by-side specs, pricing & capabilities · Updated September 2026
Price vs Intelligence
Add to comparison
2/6 models| Organization | ||
| OpenTools Score | 24 17.3 | 65 42.0 |
| Family | Gemini | Kimi |
| Status | Current | Current |
| Release Date | Jul 2026 | Nov 2025 |
| Context Window | 1.0M tokens | 262K tokens |
| Input Price | $0.30/M tokens | $0.60/M tokens |
| Output Price | $2.50/M tokens | $2.50/M tokens |
| Pricing Notes | DeepMind model page lists $0.30 input and $2.50 output per 1M tokens with no caching in the performance table. The ai.google.dev API docs were attempted through ScrapingBee and returned HTTP 400, so exact billing details beyond the official DeepMind price table were not used. | Cache read: $0.1500/M tokens |
| Capabilities | textvisionaudio-inputvideo-inputpdf-inputreasoningcodingtool-usesearch-groundingcomputer-usestructured-output | textcodeextended-thinking |
| Max Output | 64K tokens | 262K tokens |
| API Identifier | gemini-3.5-flash-lite | moonshotai/kimi-k2-thinking |
| Benchmarks | ||
| Artificial Analysis Intelligence Index | 36artificial-analysis | — |
| SWE-Bench Pro | 54.2google-deepmind | — |
| Terminal-Bench 2.1 | 54google-deepmind | — |
| GDPVal-AA v2 | 1140google-deepmind | — |
| Artificial Analysis Intelligence Index v4.3 | 22.6607artificial-analysis | — |
| GPQA Diamond | 83.8384artificial-analysis | 83.8384artificial-analysis |
| SciCode | 41.3194artificial-analysis | — |
| GPQA | — | 84.5https://huggingface.co/moonshotai/Kimi-K2-Thinking |
| MMLU-Pro | — | 84.6https://huggingface.co/moonshotai/Kimi-K2-Thinking |
| AIME 2025 | — | 94.6667artificial-analysis |
| View Gemini 3.5 Flash-Lite | View Kimi K2 Thinking | |
Cost Calculator
Enter your expected monthly token usage to compare costs.
| Model | Input | Output | Total / mo | vs Best |
|---|---|---|---|---|
| Gemini 3.5 Flash-LiteCheapest | $0.30 | $1.25 | $1.55 | — |
| Kimi K2 Thinking | $0.60 | $1.25 | $1.85 | +19% |
Gemini 3.5 Flash-Lite
Gemini 3.5 Flash-Lite is Google DeepMind general availability low-latency model for high-volume agentic systems, coding, document parsing, translation, and structured automation. Official DeepMind evidence lists 1M input tokens, 64k output tokens, text, image, video, audio, and PDF inputs, text output, tool use, Gemini API availability, and $0.30 input / $2.50 output pricing per million tokens.
Moonshot AI
Kimi K2 Thinking
Kimi K2 Thinking is a large language model from Moonshot AI. Supports up to 262,144 token context window. Available from $0.60/M input tokens.
More Comparisons
Looking for more AI models?
Browse All LLMs