PRICE~z-ai/glm-latest cached_input_per_mtok increased from 0.08 to 0.1511mPRICE~z-ai/glm-latest output_per_mtok decreased from 4.00 to 2.7911mPRICE~z-ai/glm-latest input_per_mtok decreased from 0.12 to 0.06611mPRICE~moonshotai/kimi-latest cached_input_per_mtok decreased from 0.80 to 0.4511mPRICE~moonshotai/kimi-latest input_per_mtok decreased from 0.98 to 0.9711mPRICE~deepseek/deepseek-v4-flash-latest output_per_mtok decreased from 1.04 to 0.8111mPRICE~deepseek/deepseek-v4-flash-latest input_per_mtok increased from 0.0038 to 0.008611mPRICE~deepseek/deepseek-pro-latest cached_input_per_mtok decreased from 0.10 to 0.004211mPRICE~deepseek/deepseek-pro-latest output_per_mtok decreased from 4.20 to 0.4011mPRICE~deepseek/deepseek-pro-latest input_per_mtok increased from 0.13 to 0.1311mPRICEdeepseek/deepseek-v4-flash-0731 cached_input_per_mtok increased from 0.0051 to 0.01211mPRICEdeepseek/deepseek-v4-flash-0731 input_per_mtok increased from 0.0051 to 0.01211mPRICE~moonshotai/kimi-latest cached_input_per_mtok increased from 0.29 to 0.8041mPRICE~moonshotai/kimi-latest output_per_mtok increased from 11.36 to 13.0041mPRICE~moonshotai/kimi-latest input_per_mtok increased from 0.75 to 0.9841mPRICE~deepseek/deepseek-pro-latest input_per_mtok decreased from 0.13 to 0.131hPRICE~deepseek/deepseek-pro-latest output_per_mtok increased from 0.40 to 4.201hPRICE~deepseek/deepseek-pro-latest cached_input_per_mtok increased from 0.0042 to 0.101hPRICE~moonshotai/kimi-latest input_per_mtok decreased from 0.99 to 0.751hPRICE~moonshotai/kimi-latest output_per_mtok decreased from 13.00 to 11.361hPRICE~moonshotai/kimi-latest cached_input_per_mtok decreased from 0.80 to 0.291hPRICE~moonshotai/kimi-latest input_per_mtok increased from 0.75 to 0.992hPRICE~moonshotai/kimi-latest output_per_mtok increased from 11.36 to 13.002hPRICE~moonshotai/kimi-latest cached_input_per_mtok increased from 0.29 to 0.802hPERFORMANCEGLM-5.2 → Scaleway: TTFT ↑ 20%2hPERFORMANCEDeepSeek V4 Flash → Scaleway: TTFT ↑ 35%2hPERFORMANCEMistral Medium 3.5 → Scaleway: TTFT ↓ 21%2hPERFORMANCEDeepSeek V4 Flash → Scaleway: throughput ↓ 20%2hPERFORMANCEGPT-OSS 120B → Scaleway: throughput ↑ 31%2hPERFORMANCEGPT-OSS 120B → Scaleway: TTFT ↓ 24%2hPERFORMANCEGPT-OSS 120B → Together AI: throughput ↑ 20%2hPERFORMANCEKimi K3 via OpenRouter: TTFT ↓ 49%2hPERFORMANCEGLM-5.2 via OpenRouter: reliability recovered 96.8% → 100.0%2hPERFORMANCEKimi K3 via Cortecs: throughput ↑ 184%2hPERFORMANCEKimi K3 via Cortecs: TTFT ↓ 42%2hPERFORMANCEGLM-5.2 via OpenRouter: throughput ↓ 49%2hPERFORMANCEGLM-5.2 via OpenRouter: TTFT ↑ 121%2hPERFORMANCEMiniMax M3 via OpenRouter: reliability recovered 96.8% → 100.0%2hPERFORMANCEDeepSeek V4 Pro via Cortecs: throughput ↑ 27%2hPERFORMANCEGPT-OSS 120B via OpenRouter: throughput ↓ 22%2h

Kimi K3

5 providers · MoE · 1M context · 2 regions · all models →

Kimi K3 provider ranking

#1 OverallCortecs100.0FastestCortecs2.97 sBest valueCortecsEurope #1Cortecs100.0Best toolsCortecs100.0%Best structuredCortecs100.0%
#ProviderΔ30d
01CortecsBEST OVERALLEU RESIDENT● MEASUREDAT · Routed · Kimi K3100.0%100.0%100.0%62415663.0s171P100↑ 4.6
02TogetherBEST VALUE● MEASUREDUS · Serverless · Kimi K399.5%100.0%97.6%60215144.4s104P67↑ 11.0
03OpenRouter● MEASUREDUS · Routed · Kimi K397.2%100.0%100.0%67434085.1s90P33↑ 10.3
3 of 5 providers ranked · · Methodology →
01Cortecs100.0
Best overallAT · Routed · Kimi K3
TTFT624 ms
Rel.100.0%
Tok/s171
$ Out$15.43
Tools100.0%
Δ30d+4.6
View dossier
02Together85.2
Best valueUS · Serverless · Kimi K3
TTFT602 ms
Rel.99.5%
Tok/s104
$ Out$13.50
Tools100.0%
Δ30d+11.0
View dossier
03OpenRouter74.5
US · Routed · Kimi K3
TTFT674 ms
Rel.97.2%
Tok/s90
$ Out$13.50
Tools100.0%
Δ30d+10.3
View dossier
3 of 5 providers ranked · · Methodology →
Kimi K3 inference benchmark: fastest & cheapest providers · InferenceBench