Claude Fable 5$22.000/MClaude Opus 4.8$11.000/MClaude Opus 4.7$11.000/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5.5$12.500/MGPT-5.2$5.425/MGPT-5.2-Codex$5.425/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 3.5 Pro$5.000/MGemini 3.1 Pro$5.000/MGemini 3 Pro$5.000/MGemini 2.5 Pro$3.875/MGemini 1.5 Pro$2.375/MGemini 1.0 Ultra$12.000/MGemini 1.0 Pro$0.800/MClaude Fable 5$22.000/MClaude Opus 4.8$11.000/MClaude Opus 4.7$11.000/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5.5$12.500/MGPT-5.2$5.425/MGPT-5.2-Codex$5.425/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 3.5 Pro$5.000/MGemini 3.1 Pro$5.000/MGemini 3 Pro$5.000/MGemini 2.5 Pro$3.875/MGemini 1.5 Pro$2.375/MGemini 1.0 Ultra$12.000/MGemini 1.0 Pro$0.800/M
BETA
Subscribe
← All providers·AI model pricing index
DeepInfra

DeepInfra

DeepInfra is an AI model provider.Tokenando tracks 11 DeepInfra models, with input pricing from $0.060/M and an average blended cost of $0.459/M. Its flagship model is Qwen2.5-72B (DI).

GPU serverless for popular open-weights models at aggressive prices. OpenAI-compatible API surface.

Founded 2022HQ Menlo Park, USAWebsite ↗Official pricing ↗API docs ↗
MODELS TRACKED
11
3 categories
FLAGSHIP
Qwen2.5-72B (DI)
Live API
MIN INPUT
$0.060/M
cheapest model in family
AVG BLENDED
$0.459/M
across 11 priced models
MAX CONTEXT
128K
largest window in family
Frontier
1 model
Llama 3.1 405B (DI)profile
Live API · 128K ctx
in $0.800/Mout $0.800/M
405B serverless · manual-seed
Reasoning
1 model
DeepSeek-R1 (DI)profile
Live API · 64K ctx
in $0.550/Mout $2.190/M
R1 managed · manual-seed
Efficient
9 models
Llama 3.3 70B (DI)profile
Live API · 128K ctx
in $0.230/Mout $0.400/M
Cost-effective · manual-seed
Llama 3.1 70B (DI)profile
Live API · 128K ctx
in $0.350/Mout $0.400/M
70B managed · manual-seed
Mistral 7B (DI)profile
Live API · 32K ctx
in $0.070/Mout $0.070/M
Cheapest Mistral · manual-seed
Mixtral 8x7B (DI)profile
Live API · 32K ctx
in $0.240/Mout $0.240/M
MoE managed · manual-seed
Qwen2.5-72B (DI)profile
Live API · 128K ctx
in $0.350/Mout $0.400/M
Qwen on DeepInfra · manual-seed
Gemma 2 9B (DI)profile
Live API · 8K ctx
in $0.060/Mout $0.060/M
Cheapest Gemma · manual-seed
Yi-34B-Chat (DI)profile
Live API · 4K ctx
in $0.600/Mout $0.600/M
Yi on DeepInfra · manual-seed
WizardLM-2 8x22B (DI)profile
Live API · 64K ctx
in $0.630/Mout $0.630/M
Instruction tuned · manual-seed
Phind-CodeLlama-34B (DI)profile
Live API · 16K ctx
in $0.600/Mout $0.600/M
Code model · manual-seed
Frequently Asked Questions

How many models does DeepInfra offer?

Tokenando tracks 11 DeepInfra models.

How much do DeepInfra models cost?

DeepInfra model input pricing starts at $0.060 per million tokens, with an average blended cost of $0.459 per million across the 11 priced models we track.

What is DeepInfra's flagship model?

DeepInfra's flagship model is Qwen2.5-72B (DI). It's the highest-tier DeepInfra model we track, with input pricing of $0.350 per million tokens.

What model categories does DeepInfra cover?

DeepInfra covers 3 categories: frontier, reasoning and efficient.