Tev1 4B Experimental.
1× Nvidia RTX A2000 Laptop GPU.
Most-aggressive quantisation we have a working recommendation for. Lower precision = less VRAM = cheaper hardware, at a small accuracy cost.
Cheapest hosted endpoints.
| Provider | Access | $/M in | $/M out | |
|---|---|---|---|---|
| Together AI | hosted inference | $0.042 | $0.0 | Launch ↗ |
Cheapest hardware per quantisation.
Each row is one quantisation tier (the same weights compressed differently). Lower precision → lower VRAM → cheaper hardware, at the cost of small accuracy loss. $/小时为当前最低报价。多卡配置在有同规格实例出租时使用真实实例价格,否则为单卡价格乘以卡数。
| Quantisation | Cheapest GPU config | Total VRAM | Live $/hr | tokens/sec | |
|---|---|---|---|---|---|
|
FP16
FP16 — half precision (default)
|
16 GB |
$0.019/hr
Clore.ai · 按需
|
24.0 | Compare → | |
|
FP8
FP8 — 8-bit float (Hopper / Blackwell)
|
8 GB |
$0.022/hr
Clore.ai · 按需
|
33.6 | Compare → | |
|
INT4
INT4 — 4-bit integer (~4× VRAM saving)
|
4 GB | — | 57.7 | Compare → |
What it costs per month across providers.
Estimate your monthly bill for Tev1 4B Experimental across every host that publishes per-token pricing. Slide your token volumes; the chart + table re-rank cheapest-first.
最便宜的服务商在左侧。
每月总成本 — 输入与输出 token 合计。
账单明细。
| 服务商 | 月度总计 | |
|---|---|---|
| $0.42 | 注册 ↗ |
租 GPU,而不是按 token 付费。
对于 Tev1 4B Experimental 这类开放权重模型,你可以租一块 GPU 自己提供推理服务。算法:最便宜的 GPU 租价 × 每月 730 小时。电费已包含在租用的每小时价格中。
假设 GPU 全天 24 小时运行,利用率约 85%。如果流量时高时低,API 会更省钱,GPU 则可能更贵(空闲时段仍要付租金)。盈亏平衡分析见自托管与 API 盈亏平衡工具。