BENCHMARKFirst benchmarks recorded: MiniMax M3 → Cortecs → ?5hBENCHMARKFirst benchmarks recorded: Kimi K3 → Cortecs → ?5hBENCHMARKFirst benchmarks recorded: DeepSeek V4 Pro → Cortecs → ?5hBENCHMARKFirst benchmarks recorded: GLM-4.7 → Cortecs → ?5hBENCHMARKFirst benchmarks recorded: GLM-5.2 → Cortecs → ?6h

Llama 3.3 70B on Cerebras

#4 of 10 providers for Llama 3.3 70B · #2 of 2 models on Cerebras · US
Compare with other providers
Score88.5Rel.98.9%Tools98.2%JSON98.5%TTFT110 msP95272 msTok/s2835Price$0.33 / $0.66Ctx128k4,102 runs · last 12m ago · DEV DATA

Llama 3.3 70B on other providers

ProviderScoreTTFTTok/s$ Out
Fireworks91.9231 ms251$0.48
Nebius91.8213 ms192$0.22
Groq89.2132 ms1013$0.32
CerebrasTHIS PAGE88.5110 ms2835$0.66
OpenRouter76.8284 ms216$0.48
Scaleway75.1263 ms130$0.40
Together74.4297 ms176$0.40
Nextbit72.1252 ms113$0.24
DeepInfra69.7379 ms119$0.20
OVHcloud64.0275 ms100$0.37

Other models on Cerebras

ModelScoreTTFT$ Out
Qwen3-235B97.0148 ms$1.20
Llama 3.3 70BTHIS PAGE96.2110 ms$0.66