← All providers
DeepInfra pricing
verified 2026-09-02Consistently among the cheapest hosts for open-weight models, with simple per-token serverless pricing.
data residency & retention
from DeepInfra's own policies · verified 2026-09-02Headquarters
🇺🇸 United States (US)
Where inference runs
US
Inference runs in eight secure US-based datacenters DeepInfra owns and operates ('more locations rolling out globally'); specific cities are not disclosed.
Region pinning
Enterprise plans only
No region pinning or EU endpoint for the serverless API; only the enterprise DeepCluster dedicated-cluster product offers a preferred region.
Prompt retention
Not retained
API inputs and outputs are not stored to disk and exist only in memory during the request; carve-outs for discretionary debug/security sampling, image outputs, and batch API held for a short undisclosed period.
Trains on API data
No
Submitted data is not used for training models — except when using Google or Anthropic models, where the receiving company's training policy applies.
Zero data retention
Available
De-facto ZDR is the default for content (contractual no-retain clause), but DeepInfra does not brand a formal ZDR program or toggle.
current models · $ per 1M tokens
| Model | Input /1M | Cached in | Output /1M | Context |
|---|---|---|---|---|
Llama-4-Maverick-17B-128E Llama | $0.20 | — | $0.80 | 1.0M |
DeepSeek-V4-Flash DeepSeek | $0.09 | $0.02 | $0.18 | 1.0M |
DeepSeek-V4-Flash-0731 DeepSeek | $0.08 | $0.02 | $0.18 | 1.0M |
DeepSeek-V4-Pro DeepSeek | $1.30 | $0.10 | $2.60 | 1.0M |
Qwen3.5-397B-A17B Qwen | $0.45 | $0.22 | $3 | 262K |
Llama-3.3-70B-Instruct-Turbo Llama | $0.10 | — | $0.32 | 131K |
DeepSeek-V3.2 DeepSeek | $0.26 | $0.13 | $0.38 | 164K |
Kimi-K3 Kimi | $2.85 | $0.28 | $14.25 | 1.0M |
Qwen3-Coder-480B-A35B-Instruct-Turbo Qwen | $0.30 | $0.10 | $1 | 262K |
Qwen3-235B-A22B-Instruct-2507 Qwen | $0.09 | — | $0.55 | 262K |
Kimi-K2.7-Code Kimi | $0.68 | $0.14 | $3.40 | 262K |
Kimi-K2.6 Kimi | $0.75 | $0.15 | $3.50 | 262K |
DeepInfra totals · 1M in + 1M out
DeepSeek-V4-Flash-0731
$0.2600
DeepSeek-V4-Flash
$0.2700
Llama-3.3-70B-Instruct-Turbo
$0.4200
DeepSeek-V3.2
$0.6400
Qwen3-235B-A22B-Instruct-2507
$0.6400
Llama-4-Maverick-17B-128E
$1.00
Qwen3-Coder-480B-A35B-Instruct-Turbo
$1.30
Qwen3.5-397B-A17B
$3.45
DeepSeek-V4-Pro
$3.90
Kimi-K2.7-Code
$4.08
Kimi-K2.6
$4.25
Kimi-K3
$17.10
Cost of 1M input + 1M output tokens. Bar length is square-root scaled so cheap models stay visible next to premium ones.
legacy models · still purchasable
| Model | Input /1M | Cached in | Output /1M | Context |
|---|---|---|---|---|
DeepSeek-R1-0528 DeepSeek | $0.50 | $0.35 | $2.15 | 164K |