Kimi K3 Intelligence, Performance & Price Analysis
Model Summary
Intelligence
- Rank: #4 / 189
- Score: 57 on Artificial Analysis Intelligence Index
- 4 out of 4 units for Intelligence
Speed
- Rank: #90 / 189
- Output: 62.0 tokens per second
- 2 out of 4 units for Speed
Price
- Input: $3.00 per 1M tokens
- Output: $15.00 per 1M tokens
- 3 out of 4 units for Price
Cache Hit Price
- $0.30 (-90%) USD per 1M tokens
- 3 out of 4 units for Cache Hit Price
Verbosity
- Output tokens from Intelligence Index: 130M
- 4 out of 4 units for Verbosity
Comparison Summary
Kimi K3 is amongst the leading models in intelligence, but somewhat expensive when comparing to other models of similar price. It's also slower than average and very verbose. The model supports text and image input, outputs text, and has a 1M tokens context window.
Kimi K3 scores 57 on the Artificial Analysis Intelligence Index, placing it well above average among comparable models (averaging 30). When evaluating the Intelligence Index, it generated 130M tokens, which is very verbose in comparison to the average of 63M.
Pricing for Kimi K3 is $3.00 per 1M input tokens (somewhat expensive, average: $1.75) and $15.00 per 1M output tokens (somewhat expensive, average: $8.40). In total, it cost $2690.80 to evaluate Kimi K3 on the Intelligence Index.
At 62 tokens per second, Kimi K3 is slower than average (73).
Technical Specifications
- Reasoning: Yes (this page shows the reasoning version of this model; a non-reasoning variant may also exist)
- Input modality: text, image
- Output modality: text
- Context window: 1M tokens (~1573 A4 pages of size 12 Arial font)
Intelligence Breakdown
Intelligence evaluations measured independently by Artificial Analysis (higher is better). Scores shown are from selected benchmarks:
- GDPval-AA v2: 57%
- τ³-Banking: 33%
- Terminal-Bench v2.1: 85%
- SciCode: 59%
- Humanity's Last Exam: 44%
- GPQA Diamond: 94%
- CritPt: 23%
- AA-Omniscience Accuracy: 46%
- AA-Omniscience Non-Hallucination Rate: 49%
- AA-LCR: 75%
- AA-Briefcase (Elo): 1583
- AutomationBench-AA: (score not shown individually in text)
The Artificial Analysis Intelligence Index v4.1 incorporates 9 evaluations: GDPval-AA v2, τ³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR.
