token.app โ€บ Moonshot AI โ€บ Kimi K3

Kimi K3 pricing & benchmarks

Kimi K3 is available from Moonshot AI at $3.00 per million input tokens and $15.00 per million output tokens ($6.00 blended at 3:1). It accepts up to 1,048,576 tokens of context. Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...

Pricing
Per million tokens ยท OpenRouter
Input$3.00
Output$15.00
Blended (3:1)$6.00
Specification
moonshotai/kimi-k3
Context window1.0M
Max outputโ€”
ReleasedJul 2026
LicenceOpen weights
Modalitiesin:text, in:image, out:text
Benchmarks
Independently run โ€” not vendor self-reports. Every score links its source.
BenchmarkScoreConfigRunSource
GPQA Diamond 93.1% ยฑ1.5 Max effort 2026-07-16 Independently run ยท Epoch AI
FrontierMath T1โ€“3 72.2% ยฑ2.7 Max effort 2026-07-17 Independently run ยท Epoch AI
FrontierMath T4 39.0% ยฑ7.7 Max effort 2026-07-17 Independently run ยท Epoch AI
SimpleQA Verified 42.7% ยฑ1.6 Max effort 2026-07-16 Independently run ยท Epoch AI
OTIS Mock AIME 97.2% ยฑ1.1 Max effort 2026-07-16 Independently run ยท Epoch AI
Scores published by Epoch AI, 'AI Benchmarking Hub', used under CC BY 4.0.
Cheaper models that score at least as well on GPQA Diamond
Strictly lower blended price AND an equal-or-higher GPQA Diamond score. One benchmark is one dimension โ€” a model that wins here may still be weaker at your specific task.
ModelProviderBlended $/1MGPQA Diamond
GPT-5.6 Terra OpenAI $2.25 (โˆ’63%) 93.3% Compare โ†’
Gemini 3.6 Flash Google $3.00 (โˆ’50%) 94.1% Compare โ†’
Grok 4.5 xAI $3.00 (โˆ’50%) 93.4% Compare โ†’
Gemini 3.1 Pro Preview Google $4.50 (โˆ’25%) 94.4% Compare โ†’
GPT-5.4 OpenAI $5.63 (โˆ’6%) 93.3% Compare โ†’
Where to run Kimi K3
12 options across 9 providers ยท 1.9ร— between cheapest and dearest. Blended at 3:1. โš  marks an option serving less context than the best available, or re-pricing above a prompt-length threshold. Bracketed labels are the provider's own service tier or region.
HostBlended $/1MIn / OutContextWeightsUptime 24h
Morph $5.38 $2.50 / $14.00 1.0M โ€” 98.4%
DigitalOcean $5.70 $2.85 / $14.25 1.0M โ€” 99.1%
Modal $6.00 $3.00 / $15.00 1.0M mxfp4 98.9%
BaseTen $6.00 $3.00 / $15.00 1.0M fp8 98.6%
Fireworks $6.00 $3.00 / $15.00 1.0M โ€” 98.4%
Chutes $6.00 $3.00 / $15.00 1.0M mxfp4 97.7%
Together $6.00 $3.00 / $15.00 1M โ€” 84.1%
Moonshot AI $6.00 $3.00 / $15.00 1.0M mxfp4 99.9%
Wafer $6.00 $3.00 / $15.00 1.0M fp8 99.5%
Fireworks (fast) $9.00 $4.50 / $22.50 1.0M โ€” 97.9%
Wafer (fast) $9.00 $4.50 / $22.50 1.0M fp8 98.5%
Morph (fast) $10.13 $6.00 / $22.50 1.0M โ€” 98.6%
Host prices and uptime from OpenRouter, synced 2026-08-08. Lower-precision weights (fp4, fp8) can cost less and score worse than the same model at bf16 โ€” the score above is not host-specific.
Compare Kimi K3
FAQ
How much does Kimi K3 cost?
Kimi K3 costs $3.00 per million input tokens and $15.00 per million output tokens on OpenRouter. At a 3:1 input:output mix that blends to $6.00 per million tokens.
What is Kimi K3's context window?
Kimi K3 accepts up to 1,048,576 tokens of context.
How good is Kimi K3 on benchmarks?
Kimi K3 scores 93.1% on GPQA Diamond, independently run and published by Epoch AI. 5 benchmark results are listed on this page, each with its run date and source.
Is there a cheaper model as good as Kimi K3?
Yes โ€” 5 models in our catalogue cost less per token than Kimi K3 and score at least as high on the same benchmark. The cheapest is GPT-5.6 Terra at $2.25 per million tokens blended.