DeepSeek: R1
deepseek/deepseek-r1
Cheapest provider
$0.70 / 1M
Novita
Fastest provider (p95)
100 tok/s
Azure
Intelligence (composite)
72.9
MMLU-Pro · HumanEval · math · GPQA
Per-provider performance
Latency / throughput / uptime measured across providers over the last 30 minutes of live traffic. Atlas’s router weighs these per call (with the eval-gate signal) when picking a variant for Standard and Batch tiers.
| Provider | Quant | Input $/1M | Output $/1M | Latency p50 / p95 | Throughput p50 / p95 | Uptime 30m | Success |
|---|---|---|---|---|---|---|---|
| Novita | q8· fp8 | $0.7000 | $2.5000 | 2338ms / 4336ms | 21 / 25 tok/s | 99.94% | 99.9% |
| Azure | undisclosed | $1.4850 | $5.9400 | 1251ms / 7873ms | 76 / 100 tok/s | 100.00% | 100.0% |
“—” 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 breakdown
Composite score is a weighted average of public benchmarks (30% MMLU-Pro, 25% code pass@1, 25% math, 20% GPQA). Numbers come from model cards and the Artificial Analysis intelligence harness; missing components are renormalised over what’s present.
MMLU-Pro
84.0
broad reasoning
Code
63.5
pass@1 (HumanEval / LiveCodeBench)
AIME 2025
70.0
math accuracy
GPQA Diamond
71.5
hard reasoning
Source: DeepSeek-R1 model card (MMLU-Pro 84.0, GPQA-Diamond 71.5, LiveCodeBench 63.5, AIME25 70.0); huggingface.co/deepseek-ai/DeepSeek-R1-0528
How Atlas routes DeepSeek: R1
- Realtime tier — direct passthrough at the upstream’s native precision. Best for hard latency / quality guarantees.
- Standard tier — Atlas picks the cheapest provider variant whose quantization has stayed green on your operation’s eval gates. For DeepSeek: R1 that’s currently Novita at $0.70/1M.
- Batch tier — async, biggest discount. Roadmapped to use provider batch APIs (OpenAI / Anthropic) where available and queued spot capacity for open-weight workloads.