DeepInfra
InferenceCheap serverless inference — Llama, Mistral, FLUX at low cost
Operational
All systems responding normally
Last checked 30/06/2026, 4:28:36 pm
585ms response
Uptime History100.00% uptime
2026-06-24Today
Uptime
100.00%
Avg Latency
735ms
P95 Latency
1024ms
Fastest
471ms
Checks
150
Response Time
Last 60 checks471ms min735ms avg1640ms max
💰 Pricing
llama-3.3-70b
Input: $0.23/1MOutput: $0.4/1M
Cheapest 70B inference. meta-llama-3.1-8b: $0.06/$0.06
⚡ Rate Limits
free
RPM: 30TPM: 120,000
Free: 30 RPM, 120K TPM. Paid: up to 1000 RPM.
🤖 Models (1)
| Model | Task | Context | Vision | Tools | JSON |
|---|---|---|---|---|---|
| Llama 3.3 70B Cheap inference. $0.23/$0.40 per 1M tokens. | llm | 128k | — | ✅ | ✅ |
Recent Checks
Showing last 15Operational
585ms30 June, 04:28 pm
Operational
609ms30 June, 02:52 pm
Operational
661ms30 June, 01:22 pm
Operational
754ms30 June, 12:06 pm
Operational
798ms30 June, 11:16 am
Operational
785ms30 June, 10:08 am
Operational
658ms30 June, 09:39 am
Operational
751ms30 June, 09:12 am
Operational
818ms30 June, 08:41 am
Operational
530ms30 June, 08:18 am
Operational
669ms30 June, 07:48 am
Operational
761ms30 June, 07:11 am
Operational
620ms30 June, 06:30 am
Operational
705ms30 June, 05:48 am
Operational
705ms30 June, 05:02 am
Other Inference Providers
Groq
LPU inference — fastest tokens per second on the market
Cerebras
Wafer-scale chip inference — 1,000+ tokens/sec
Together AI
Open-source model inference — Llama, Mixtral, FLUX
Fireworks AI
Fast open-model inference — FireFunction, Llama, Mixtral
OpenRouter
Unified API across 200+ models — route by price or speed
Hugging Face
Serverless inference API — 100k+ open models on demand