NEWMEN
← back to all models

DeepSeek: DeepSeek V3 0324

deepseek/deepseek-chat-v3-0324

DeepSeekopen-weight164K context6 providersIntelligence 71.0· est.

Cheapest provider

$0.20 / 1M

DeepInfra

Fastest provider (p95)

76 tok/s

DeepInfra

Intelligence (estimated)

71.0

Family + generation + popularity + price tier

Per-provider performance

Latency / throughput / uptime / price measured across providers over the last 30 minutes of live traffic. This is what proves “sourced cheapest” — Atlas mode draws on these per call to serve the cheapest path that holds quality.

ProviderQuantInput $/1MOutput $/1MLatency p50 / p95Throughput p50 / p95Uptime 30mSuccess
DeepInfraq4· fp4$0.2000$0.7700420ms / 1319ms34 / 76 tok/s98.43%99.7%
AtlasCloudq8· fp8$0.2160$0.88001470ms / 3799ms5 / 16 tok/s99.43%99.5%
ModelRunq4· fp4$0.2200$0.80001201ms / 3767ms35 / 49 tok/s99.81%98.9%
SiliconFlowq8· fp8$0.2500$1.00001482ms / 3829ms18 / 27 tok/s98.23%98.6%
Novitaq8· fp8$0.2700$1.12001386ms / 3685ms23 / 31 tok/s99.07%99.0%
GMICloudq8· fp8$0.2900$1.14002698ms / 6040ms17 / 31 tok/s98.95%98.6%

“—” means live telemetry hasn’t accumulated enough recent traffic for that endpoint. “undisclosed” means the provider serves the model but doesn’t expose the quantization label (typically running fp8 / int8 internally).

Intelligence estimate

No public benchmark numbers indexed for this model yet, so the leaderboard score is derived from the catalogue data we sync: model family standing, generation, and live usage rank — applied identically to open- and closed-weight models. The heuristic ceiling sits below confirmed-benchmark frontier models so curated rankings stay clearly on top.

Want a curated score? File a benchmark report and we’ll add it to the next sync.

How Atlas mode sources DeepSeek: DeepSeek V3 0324

  • Strict mode — pin DeepSeek: DeepSeek V3 0324 exactly and we pass it straight through, sourced from the cheapest provider above. The same model, no substitutions — currently DeepInfra at $0.20/1M.
  • Atlas mode — the default. Each call is auto-optimized for the cheapest path that holds quality, at least 5% off going direct from call one and climbing as it ramps. You always see which model served the call and exactly what you saved — thumbs-down anything you don’t like for a full refund.