PERFORMANCEGLM-5.2 → Scaleway: TTFT ↑ 20%1mPERFORMANCEDeepSeek V4 Flash → Scaleway: TTFT ↑ 35%1mPERFORMANCEMistral Medium 3.5 → Scaleway: TTFT ↓ 21%1mPERFORMANCEDeepSeek V4 Flash → Scaleway: throughput ↓ 20%1mPERFORMANCEGPT-OSS 120B → Scaleway: throughput ↑ 31%1mPERFORMANCEGPT-OSS 120B → Scaleway: TTFT ↓ 24%1mPERFORMANCEGPT-OSS 120B → Together AI: throughput ↑ 20%1mPERFORMANCEKimi K3 via OpenRouter: TTFT ↓ 49%1mPERFORMANCEGLM-5.2 via OpenRouter: reliability recovered 96.8% → 100.0%1mPERFORMANCEKimi K3 via Cortecs: throughput ↑ 184%1mPERFORMANCEKimi K3 via Cortecs: TTFT ↓ 42%1mPERFORMANCEGLM-5.2 via OpenRouter: throughput ↓ 49%1mPERFORMANCEGLM-5.2 via OpenRouter: TTFT ↑ 121%1mPERFORMANCEMiniMax M3 via OpenRouter: reliability recovered 96.8% → 100.0%1mPERFORMANCEDeepSeek V4 Pro via Cortecs: throughput ↑ 27%1mPERFORMANCEGPT-OSS 120B via OpenRouter: throughput ↓ 22%1mPERFORMANCEGLM-4.7 via OpenRouter: TTFT ↓ 78%1mPERFORMANCEGPT-OSS 20B → Groq: TTFT ↑ 24%1mPERFORMANCEKimi K3 → Together AI: TTFT ↓ 34%1mPERFORMANCEDeepSeek V4 Flash via OpenRouter: throughput ↓ 17%1mPERFORMANCELlama 3.3 70B via Cortecs: TTFT ↓ 51%1mPERFORMANCEQwen3 235B via OpenRouter: throughput ↑ 37%1mPERFORMANCEQwen3 235B via OpenRouter: TTFT ↓ 29%1mPERFORMANCEDeepSeek V4 Flash → Together AI: throughput ↑ 19%1mPERFORMANCEGLM-5.3 via OpenRouter: throughput ↓ 17%1mPERFORMANCEGPT-OSS 20B via Cortecs: throughput ↑ 28%1mPERFORMANCELlama 3.3 70B → Together AI: TTFT ↓ 33%1mPRICE~moonshotai/kimi-latest cached_input_per_mtok decreased from 0.80 to 0.292mPRICE~moonshotai/kimi-latest output_per_mtok decreased from 13.00 to 11.362mPRICE~moonshotai/kimi-latest input_per_mtok decreased from 0.99 to 0.752mPRICE~deepseek/deepseek-v4-flash-latest cached_input_per_mtok decreased from 0.0077 to 0.00112mPRICE~deepseek/deepseek-v4-flash-latest output_per_mtok decreased from 1.28 to 1.042mPRICE~deepseek/deepseek-v4-flash-latest input_per_mtok decreased from 0.0077 to 0.00382mPRICEz-ai/glm-5.1 cached_input_per_mtok increased from 0.18 to 0.262mPRICEz-ai/glm-5.1 output_per_mtok increased from 3.03 to 4.402mPRICEz-ai/glm-5.1 input_per_mtok increased from 0.96 to 1.402mPRICEtencent/hy4-preview cached_input_per_mtok increased from 0.038 to 0.0422mPRICEtencent/hy4-preview input_per_mtok increased from 0.75 to 0.832mPRICEtencent/hy3 cached_input_per_mtok increased from 0.021 to 0.0332mPRICEtencent/hy3 output_per_mtok increased from 0.33 to 0.532m

GPT-OSS 120B on Cerebras

#2 of 6 providers for GPT-OSS 120B · #1 of 1 models on Cerebras · US
Compare with other providers
Score99.5Rel.100.0%Tools100.0%JSON100.0%TTFT184 msP95323 msTok/s1749Price$0.35 / $0.75Ctx128k217 results · last 1135m ago

GPT-OSS 120B on other providers

ProviderScoreTTFTTok/s$ Out
Groq99.7264 ms486$0.60
CerebrasTHIS PAGE99.5184 ms1749$0.75
Cortecs99.4240 ms358$0.46
Scaleway94.4179 ms96$0.69
OpenRouter88.8474 ms68$0.17
Together72.5182 ms138$0.60

Other models on Cerebras

ModelScoreTTFT$ Out
GPT-OSS 120BTHIS PAGE91.2184 ms$0.75

3 more models declared in Cerebras's catalogue — not yet measured from the canonical vantage.