Kimi K2 7B: pricing, context window and benchmarks
Kimi K2 7B is a small open-weight model from Moonshot AI. It offers a 16,384-token context window, costs $1.72 per million input tokens and $8.97 per million output tokens, and averages 62.7 across the index's benchmark suite.
Pricing
Kimi K2 7B costs $1.72 per million input tokens and $8.97 per million output tokens. A blended 3:1 input-to-output mix works out to about $3.53 per million tokens. Cached input is listed at $0.24 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $1.72 | per 1M tokens |
| Output | $8.97 | per 1M tokens |
| Cached input | $0.24 | per 1M tokens |
| Blended 3:1 | $3.53 | estimate per 1M tokens |
| 1M in + 1M out | $10.69 | estimate |
Benchmarks
Scores below are modelled estimates on a 0–100 scale. Use them to compare relative position, not as official vendor results.
| Benchmark | Score | Rank | Percentile | Relative |
|---|---|---|---|---|
| MMLU-Pro | 66.5 | 431 | 28th | |
| GPQA Diamond | 57.0 | 440 | 27th | |
| SWE-bench Verified | 52.5 | 401 | 33th | |
| HumanEval | 67.5 | 416 | 31th | |
| MATH-500 | 64.6 | 421 | 30th | |
| AIME 2025 | 51.1 | 450 | 25th | |
| LiveCodeBench | 50.4 | 444 | 26th | |
| IFEval | 75.0 | 398 | 34th | |
| TruthfulQA | 67.9 | 414 | 31th | |
| HellaSwag | 78.1 | 400 | 33th | |
| MMMU | 59.2 | 437 | 27th | |
| Arena Elo | 1304.4 | 412 | 31th |
What Kimi K2 7B is good for
Kimi K2 7B is a small model. That makes it a good fit for classification, extraction, routing, summarization and other high-volume jobs where latency and cost dominate. It is not the right pick for open-ended frontier reasoning, but at $1.72 per million input tokens it can handle a large share of production traffic.
Strengths
- Input price of $1.72 per million tokens is at or below the index median.
- Open weights make self-hosting, fine-tuning and data-residency review possible.
- Accepts more than text, including image input where the provider supports it.
Trade-offs
- Context window of 16,384 tokens is small for long-document work.
- Average benchmark score is below the index median; test it on your own task.
- Less reliable on open-ended reasoning and complex multi-step instructions.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| DeepSeek-R1 7B | DeepSeek | 32768 | $1.79 | 63.0 |
| Zephyr | Hugging Face | 8192 | $1.61 | 62.8 |
| Hunyuan-Turbo Nano | Tencent | 32768 | $4.65 | 62.7 |
| Kimi K2 7B | Moonshot AI | 16384 | $1.72 | 62.7 |
| GPT-NeoX Instruct | EleutherAI | 16384 | $1.52 | 62.6 |
| Zephyr Instruct | Hugging Face | 4096 | $1.66 | 62.4 |
FAQ
How much does Kimi K2 7B cost?
Kimi K2 7B is listed at $1.72 per million input tokens and $8.97 per million output tokens. Cached input is $0.24 per million tokens. A blended 3:1 input-to-output mix is about $3.53 per million tokens.
What is the context window of Kimi K2 7B?
Kimi K2 7B has a context window of 16,384 tokens and can emit up to 16,384 output tokens per request in this index.
Is Kimi K2 7B open weights?
Yes. Kimi K2 7B is listed as an open-weight model under a open weights licence.
How does Kimi K2 7B score on benchmarks?
Kimi K2 7B averages 62.7 across the scored benchmarks in this index, ranking #423 of 600 models. Its strongest relative result is Arena Elo at 1304.4.
What is Kimi K2 7B best for?
Kimi K2 7B is a small model. That makes it a good fit for classification, extraction, routing, summarization and other high-volume jobs where latency and cost dominate. It is not the right pick for open-ended frontier reasoning, but at $1.72 per million input tokens it can handle a large share of production traffic.