DeepSeek: DeepSeek V3.2
deepseek/deepseek-v3.2
Cheapest provider
$0.23 / 1M
StreamLake
Fastest provider (p95)
—
No throughput data yet — populated as traffic accumulates
Intelligence (composite)
83.3
MMLU-Pro · HumanEval · math · GPQA
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.
| Provider | Quant | Input $/1M | Output $/1M | Latency p50 / p95 | Throughput p50 / p95 | Uptime 30m | Success |
|---|---|---|---|---|---|---|---|
| StreamLake | undisclosed | $0.2288 | $0.3432 | — | — | 99.17% | — |
| Baidu | q8· fp8 | $0.2520 | $0.3780 | — | — | 99.77% | — |
| SiliconFlow | q8· fp8 | $0.2590 | $0.4200 | — | — | 99.74% | — |
| AtlasCloud | q8· fp8 | $0.2600 | $0.3800 | — | — | 99.48% | — |
| DeepInfra | q4· fp4 | $0.2600 | $0.3800 | — | — | 98.49% | — |
| Novita | q8· fp8 | $0.2690 | $0.4000 | — | — | 99.92% | — |
| Parasail | q8· fp8 | $0.2800 | $0.4500 | — | — | 98.34% | — |
| Alibaba | undisclosed | $0.3705 | $1.1115 | — | — | 99.79% | — |
| DigitalOcean | undisclosed | $0.4250 | $1.3600 | — | — | 99.83% | — |
| Friendli | undisclosed | $0.5000 | $1.5000 | — | — | 99.97% | — |
| undisclosed | $0.5600 | $1.6800 | — | — | 99.50% | — | |
| SambaNova | undisclosed | $3.0000 | $4.5000 | — | — | — | — |
“—” 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
85.0
broad reasoning
Code
79.0
pass@1 (HumanEval / LiveCodeBench)
AIME 2025
88.0
math accuracy
GPQA Diamond
80.0
hard reasoning
Source: Artificial Analysis (AA-aligned) via MiniMax-M2 table — DeepSeek-V3.2 (MMLU-Pro 85, GPQA-D 80, LiveCodeBench 79, AIME25 88); github.com/MiniMax-AI/MiniMax-M2
How Atlas mode sources DeepSeek: DeepSeek V3.2
- Strict mode — pin DeepSeek: DeepSeek V3.2 exactly and we pass it straight through, sourced from the cheapest provider above. The same model, no substitutions — currently StreamLake at $0.23/1M.
- 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.