Inference host

DeepInfra

LLM prices per million tokens

deepinfra.com

Official pricing

Models: 57
Routes: 61
Cheapest: $0.01; Qwen3 Embedding 8B input
Batch: n/a; not stated

Models on DeepInfra

Qwen3 Embedding 8BDeepInfra, us$0.01 / Free
Qwen3 Embedding 4BDeepInfra, us$0.02 / Free
Mistral NemoDeepInfra, fp8, fp8$0.019 / $0.03+4% vs headline
Llama 3.1 8B InstructDeepInfra, fp8, fp8$0.02 / $0.04
gpt-oss-20bDeepInfra, bf16, bf16$0.03 / $0.14+60% vs headline
Mistral Small 3DeepInfra, fp8, fp8$0.05 / $0.08
Gemma 3 4BDeepInfra, bf16, bf16$0.05 / $0.1
gpt-oss-120bDeepInfra, bf16, bf16$0.037 / $0.17+8% vs headline
Gemma 3 12BDeepInfra, bf16, bf16$0.05 / $0.15
DeepSeek V4 Flash 0731DeepInfra, fp8, fp8$0.06 / $0.18+14% vs headline
Gemma 3 27BDeepInfra, fp8, fp8$0.08 / $0.16
Mistral Small 3.2 24BDeepInfra, fp8, fp8$0.075 / $0.2
DeepSeek V4 Flash 0423DeepInfra, fp8, fp8$0.09 / $0.18+2% vs headline
Qwen3.5-9BDeepInfra, bf16, bf16$0.1 / $0.15+22% vs headline
GLM 5.3 FlashDeepInfra, fp4, fp4$0.075 / $0.25+73% vs headline
Qwen3 32BDeepInfra, fp8, fp8$0.08 / $0.28
Gemma 4 26B A4BDeepInfra, fp8, fp8$0.07 / $0.34+59% vs headline
Qwen3 14BDeepInfra, fp8, fp8$0.12 / $0.24-62% vs headline
Llama 4 ScoutDeepInfra, fp8, fp8$0.1 / $0.3
Gemma 4 31BDeepInfra, turbo, fp4$0.09 / $0.34
Llama 3.3 70B InstructDeepInfra, turbo, fp8$0.1 / $0.32
Llama Guard 4 12BDeepInfra, bf16, bf16$0.18 / $0.18
Gemma 4 31BDeepInfra, fp8, fp8$0.13 / $0.38+26% vs headline
Qwen3 235B A22B Instruct 2507DeepInfra, fp8, fp8$0.09 / $0.55-22% vs headline
DeepSeek V4.1 FlashDeepInfra, fp8, fp8$0.14 / $0.42-20% vs headline
Qwen3 30B A3BDeepInfra, fp8, fp8$0.12 / $0.5-5% vs headline
Qwen3 VL 30B A3B InstructDeepInfra, fp8, fp8$0.15 / $0.6+15% vs headline
gpt-oss-120bDeepInfra, turbo, bf16$0.15 / $0.6+304% vs headline
DeepSeek V3.2DeepInfra, fp4, fp4$0.26 / $0.38+24% vs headline
Qwen3.6 35B A3BDeepInfra, fp8, fp8$0.1 / $0.95+47% vs headline
DeepSeek V4 Flash Vision ExpDeepInfra, fp8, fp8$0.216 / $0.647
Qwen3 Next 80B A3B InstructDeepInfra, fp8, fp8$0.09 / $1.1+28% vs headline
Llama 4 MaverickDeepInfra, base, fp8$0.2 / $0.8+15% vs headline
Qwen3.5-35B-A3BDeepInfra, fp8, fp8$0.14 / $1-21% vs headline
Qwen3 VL 235B A22B InstructDeepInfra, fp8, fp8$0.2 / $0.88-19% vs headline
Qwen2.5 72B InstructDeepInfra, fp8, fp8$0.36 / $0.4
gpt-oss-120bDeepInfra, fp8, fp8$0.2 / $0.95+496% vs headline
Gemma 4 31BDeepInfra, ultra, fp8$0.27 / $0.76+157% vs headline
Llama 3.1 70B InstructDeepInfra, turbo, fp8$0.4 / $0.4
DeepSeek V3 0324DeepInfra, fp4, fp4$0.24 / $0.9
DeepSeek V3.1DeepInfra, fp4, fp4$0.25 / $0.95
DeepSeek V3DeepInfra, fp4, fp4$0.32 / $0.89+3% vs headline
Qwen3 Coder 480B A35BDeepInfra, turbo, fp4$0.3 / $1
Muse Glimmer 30BDeepInfra, bf16, bf16$0.3 / $1.2+5% vs headline
Qwen3.8 27BDeepInfra, bf16, bf16$0.15 / $1.88-39% vs headline
GLM 4.7DeepInfra, fp4, fp4$0.4 / $1.75
Qwen3.5-122B-A10BDeepInfra, fp4, fp4$0.29 / $2.4+14% vs headline
Qwen3.5-27BDeepInfra, fp8, fp8$0.26 / $2.6+58% vs headline
GLM 5.2DeepInfra, fp4, fp4$0.563 / $1.8
GLM 4.6DeepInfra, fp4, fp4$0.5 / $2+15% vs headline
R1 0528DeepInfra, fp4, fp4$0.5 / $2.15
Qwen3.6 27BDeepInfra, fp8, fp8$0.32 / $3.2+3% vs headline
GLM 5.3DeepInfra, fp4, fp4$0.563 / $2.5
Qwen3.5 397B A17BDeepInfra, fp8, fp8$0.45 / $3+24% vs headline
Kimi K2.7 CodeDeepInfra, fp4, fp4$0.68 / $3.4+6% vs headline
Kimi K2.6DeepInfra, fp4, fp4$0.75 / $3.5+40% vs headline
DeepSeek V4 Pro 0813DeepInfra, fp8, fp8$1.3 / $2.6+64% vs headline
DeepSeek V4 Pro 0423DeepInfra, fp8, fp8$1.3 / $2.6+38% vs headline
GLM 5.1DeepInfra, fp4, fp4$1.05 / $3.5+12% vs headline
Qwen3.8 2.4T A95BDeepInfra, fp4, fp4$2 / $6
Kimi K3DeepInfra, bf16, bf16$2.85 / $14.3+53% vs headline

Notes from the pricing page

Caching: Cached input prices listed per model (e.g. $0.06 / $0.015 cached)

Across fru.devCompany profile

List prices from providers' pages on the date shown; your contract may differ. Logos via logo.dev; trademarks belong to their owners.