Inference cloud
How DeepInfra compares
unverifiedcatalog v1
Serverless per-token rates and dedicated GPU-hour rates.
Regions and compliance
Regions: not published / unverified
Compliance: not published / unverified
Token offerings
DeepSeek-V4-Flash
serverless_api · DeepSeek-V4-Flash · open-weight not published / unverified
- Input / 1M
- 0.09deepinfra.comchecked 2026-10-20“DeepSeek-V4-Flash | 1024k | $0.09 / $0.018 cached | $0.18”
- Cached input / 1M
- 0.018deepinfra.comchecked 2026-10-20“DeepSeek-V4-Flash | $0.018 cached”
- Output / 1M
- 0.18deepinfra.comchecked 2026-10-20“DeepSeek-V4-Flash | $0.18”
Context 1,024,000 · tokens/s not published · TTFT not published
Llama-3.3-70B-Instruct-Turbo
serverless_api · Llama-3.3-70B-Instruct-Turbo · open-weight not published / unverified
- Input / 1M
- 0.1deepinfra.comchecked 2026-10-20“Llama-3.3-70B-Instruct-Turbo | 128k | $0.10 | $0.32”
- Cached input / 1M
- not published / unverifieddeepinfra.comchecked 2026-10-20“Cached input price not published.”
- Output / 1M
- 0.32deepinfra.comchecked 2026-10-20“Llama-3.3-70B-Instruct-Turbo | $0.32”
Context 128,000 · tokens/s not published · TTFT not published
Compute SKUs
Dedicated H100 80GB 1× H100 · gpu hour · in the run-cost model
On demand
2.9deepinfra.comchecked 2026-10-20“H100 | 80GB | $2.90 / GPU-hour”Reserved
not published / unverifieddeepinfra.comchecked 2026-10-20“Dedicated clusters say contact dedicated@deepinfra.com. No reserved number.”Spot
not published / unverifieddeepinfra.comchecked 2026-10-20“No spot rate was listed.”Custom LLM uptime price. Billed in minute granularity.
Dedicated H200 141GB 1× H200 · gpu hour · shown raw, not in C_1k
On demand
3.49deepinfra.comchecked 2026-10-20“H200 | 141GB | $3.49 / GPU-hour”Reserved
not published / unverifieddeepinfra.comchecked 2026-10-20“No reserved number was listed.”Spot
not published / unverifieddeepinfra.comchecked 2026-10-20“No spot rate was listed.”