The board shows this model's cheapest measured offer ($0.233/1M); the smaller figure under it is the mean of the 23 offers listed below ($0.447/1M). Add those up and divide — it matches. 2 offer(s) above 10x the median were left out of the average and flagged ⚠ in the table below — they are still listed, and still compete for the cheapest price. 24h token volume (64B) is sourced from OpenRouter public rankings — today's only public source of absolute per-model token counts (Vercel AI Gateway publishes share-of-traffic percentages); as more sources come online, the per-router breakdown will sum to this total here.
| Provider | Router / Source | $ in /1M | $ out /1M | TTFT | TPS | Uptime | Evidence |
|---|---|---|---|---|---|---|---|
| Baidu (fp8) | OpenRouter | $0.207 | $0.311 | 1260ms | 29 | 99.8% | API-measured |
| StreamLake (fp8) | OpenRouter | $0.214 | $0.322 | 1388ms | 22 | 99.9% | API-measured |
| deepinfra | HuggingFace Router | $0.260 | $0.380 | 724ms | 31 | — | API-measured |
| DeepInfra (fp4) | OpenRouter | $0.260 | $0.380 | 1334ms | 4 | 71.1% | API-measured |
| AtlasCloud (fp8) | OpenRouter | $0.260 | $0.380 | 1143ms | 23 | 99.9% | API-measured |
| SiliconFlow (fp8) | OpenRouter | $0.259 | $0.420 | 1798ms | 22 | 99.9% | API-measured |
| novita | Pricebook (LiteLLM) | $0.269 | $0.400 | — | — | — | Official list |
| Novita (fp8) | OpenRouter | $0.269 | $0.400 | 1215ms | 20 | 100.0% | API-measured |
| novita | HuggingFace Router | $0.269 | $0.400 | 1480ms | 34 | — | API-measured |
| novita-ai | Novita AI | $0.269 | $0.400 | — | — | — | API-measured |
| requesty | Requesty | $0.269 | $0.400 | — | — | — | API-measured |
| deepseek | Pricebook (LiteLLM) | $0.280 | $0.400 | — | — | — | Official list |
| gmi | Pricebook (LiteLLM) | $0.280 | $0.400 | — | — | — | Official list |
| openrouter | Pricebook (LiteLLM) | $0.280 | $0.400 | — | — | — | Official list |
| GMICloud (fp8) | OpenRouter | $0.290 | $0.430 | 2328ms | 28 | 99.5% | API-measured |
| Venice | OpenRouter | $0.330 | $0.480 | 3615ms | 7 | 4.0% | API-measured |
| Alibaba (fp8) | OpenRouter | $0.370 | $1.11 | 1009ms | 40 | 96.9% | API-measured |
| DigitalOcean | OpenRouter | $0.425 | $1.36 | 1014ms | 18 | 83.8% | API-measured |
| Friendli | OpenRouter | $0.500 | $1.50 | 230ms | 72 | 100.0% | API-measured |
| OpenRouter | $0.560 | $1.68 | 1891ms | 18 | 99.4% | API-measured | |
| azure_ai | Pricebook (LiteLLM) | $0.580 | $1.68 | — | — | — | Official list |
| burncloud | burncloud | $0.580 | $1.68 | — | — | — | API-measured |
| Phala | OpenRouter | $1.00 | $1.00 | 4160ms | 7 | 99.7% | API-measured |
| sambanova ⚠ outlier | Pricebook (LiteLLM) | $3.00 | $4.50 | — | — | — | Official list |
| SambaNova ⚠ outlier | OpenRouter | $3.00 | $4.50 | 2212ms | 25 | 99.5% | API-measured |
Volume trend
Last 30 days: 62B → 64B (+3%); range 48B–65B tokens/day. Source: OpenRouter public rankings; refreshed hourly.
Data & method
Cheapest ranked by blended = (3×in + out) ÷ 4; performance columns show only when a source provides them, never estimated. “Official list” = LiteLLM open pricebook (MIT); “API-measured” = provider/aggregator public APIs. Machine-readable data: /api/v1/models/deepseek-v3.2.json. Free reuse requires attribution to tkx.org.
About
DeepSeek V3.2 is an AI model, a machine-learning system designed for natural language processing tasks. It is served by multiple inference providers on AI model marketplaces and is listed under the organization tag "deepseek".
AI-generated summary from public information — is this yours?
FAQ
How much does DeepSeek V3.2 cost?
Cheapest measured offer right now: $0.2072/1M input, $0.3108/1M output via Baidu (fp8) — across 19 tracked providers, refreshed hourly.
Who serves DeepSeek V3.2?
19 providers currently serve it. The table above lists each with input/output price, TTFT, TPS, uptime and an evidence tier.
How heavily is DeepSeek V3.2 used?
DeepSeek V3.2 routed 64B tokens in the last 24h. The daily bars above show routed volume over recent days. Source: OpenRouter public rankings — one of the few routers publishing per-model usage; TKX refreshes hourly.