FriendliAI
- Categories
- Inference hosts
- Owned by
- FriendliAI Corp. (private)
- HQ
- US
- Website
- https://friendli.ai
- Pricing page
- https://friendli.ai/pricing
- Fee model
- per_token+per_second_gpu
- Fees
- Model APIs per token; Dedicated Endpoints per GPU-hour billed per second: A100 80GB $2.9/h, H100 $3.9/h, H200 $4.5/h, B200 $8.9/h, B300 $12.0/h through Sep 30 2026, rising Oct 1 to A100 $4.0, H100 $5.0, H200 $7.0, B200 $9.0, B300 $12.0; BYOG (own GPUs) custom
- Free tier
- no
- OpenAI-compatible API
- yes
- Bring your own key
- no
- Notes
- HQ San Francisco with Seoul hub; SOC 2 Type II, HIPAA. whisper-large-v3 STT $0.0015/audio minute. Cached-input discounts for non-listed models via enterprise.
- Sources
- https://friendli.ai/pricing collected 2026-09-18
LLM prices (7 models, USD per 1M tokens)
| Model | Input | Output | Cached input | Context | Source |
|---|---|---|---|---|---|
| DeepSeek V3.2 deepseek-ai/DeepSeek-V3.2 |
$0.50 | $1.50 | $0.25 | 164k | ✓ src |
| Gemma 4 31B google/gemma-4-31B-it |
$0.14 | $0.40 | — | 262k | ✓ src |
| GLM 5.1 zai-org/GLM-5.1 |
$1.40 | $4.40 | $0.26 | 203k | ✓ src |
| GLM-5.2 zai-org/GLM-5.2 |
$1.40 | $4.40 | $0.26 | 1.05M | ✓ src |
| GLM-5.3 zai-org/GLM-5.3 |
$1.40 | $4.40 | $0.26 | 1.05M | ✓ src |
| GLM-5.3 Flash zai-org/GLM-5.3-Flash |
$0.15 | $0.50 | $0.03 | 1.05M | ✓ src |
| MiniMax M2.5 MiniMaxAI/MiniMax-M2.5 |
$0.30 | $1.20 | $0.06 | 197k | ✓ src |