Prime Inference
Inference hostPrime Inference runs open and licensed models on its own hardware and sells them by usage, often faster or cheaper than the maker.
Models122
- Cheapest for
- 0 models
- Serves
- 1 model types
Visits in September 2026172K−41.3%Unique visitors 70K
- Apr
- May
- Jun
- Jul
- Aug
- Sep
Models by type
Models on Prime Inference
| Model | Input | Output | vs official | Context |
|---|---|---|---|---|
 Llama 3.2 3B InstructMeta | $0.051/1M | $0.335/1M from $0.05 at DevPass (LLM Gateway) | +570% | — |
 MiniMax-M2.7MiniMax | $0.60/1M | $2.40/1M from $0.60 at EmpirioLabs AI | +300% | — |
 Mistral Large 3Mistral | $0.50/1M | $1.50/1M | ±0% | — |
| Mixtral 8x22B InstructMistral | $2.20/1M | $6.60/1M from $6 at OpenRouter | +10% | — |
| OpenAI GPT-4.1OpenAI | $2/1M | $8/1M from $6.40 at APIMart | ±0% | — |
| OpenAI GPT-4o miniOpenAI | $0.15/1M | $0.60/1M from $0.48 at APIMart | ±0% | — |
| OpenAI GPT-5 MiniOpenAI | $0.25/1M | $2/1M from $1.60 at APIMart | ±0% | — |
| OpenAI GPT-5 NanoOpenAI | $0.05/1M | $0.40/1M from $0.32 at APIMart | ±0% | — |
| OpenAI GPT-5.2OpenAI | $1.75/1M | $14/1M from $11.20 at APIMart | ±0% | — |
| OpenAI GPT-5.3 CodexOpenAI | $1.75/1M | $14/1M from $11.20 at APIMart | +25% | — |
| OpenAI GPT-5.4OpenAI | $2.50/1M | $15/1M from $12 at APIMart | ±0% | — |
| OpenAI GPT-5.4 MiniOpenAI | $0.75/1M | $4.50/1M from $3.60 at APIMart | ±0% | — |
| OpenAI GPT-5.4 NanoOpenAI | $0.20/1M | $1.25/1M from $1 at APIMart | ±0% | — |
| OpenAI GPT-5.4 ProOpenAI | $30/1M | $180/1M from $144 at APIMart | ±0% | — |
| OpenAI GPT-5.5OpenAI | $5/1M | $30/1M from $24 at APIMart | ±0% | — |
 Qwen3 235B A22B Instruct 2507Alibaba | $0.25/1M | $1/1M from $0.10 at Requesty | +900% | — |
| Qwen3 235B A22b Thinking 2507Alibaba | $0.30/1M | $3/1M from $2.30 at Hugging Face | +30% | — |
| Qwen3 30B-A3B Instruct 2507Alibaba | $0.13/1M | $0.52/1M from $0.30 at Cortecs | +73% | — |
| Qwen3 VL 235B A22B InstructAlibaba | $0.30/1M | $1.50/1M from $0.88 at Deep Infra | +70% | — |
| Qwen3 VL 30B A3B InstructAlibaba | $0.20/1M | From$0.70/1M from $0.389 at SiliconFlow | +80% | — |
| Qwen3.5 FlashAlibaba | $0.15/1M | $0.60/1M from $0.26 at OpenRouter | +109% | — |
| Qwen3.7 FlashAlibaba | $0.03/1M | $0.13/1M from $0.091 at APIMart | +18% | — |
Prime Inference at a glance
122 models in total: 122 text. The cheapest is Qwen3.5 0.8B at $0.08 /1M. GPT-6 Astra has the largest context window, 1.05M tokens. The most searched is Gemini 4 Argon, with about 2M Google searches a month. The newest is Kimi K3.1 Pro Highspeed, released Oct 23, 2026.
Cheapest
- Qwen3.5 0.8B$0.08/1M
Laguna XS 2.1$0.12/1M- Qwen3.7 Flash$0.13/1M
- Mistral Nemo$0.17/1M
- Laguna S 2.1$0.18/1M
Largest context
- GPT-6 Astra1.05M
- GPT-6.1 Sol1.05M
- GPT-6 Sol1.05M
- GPT-6 Luna1.05M
- GPT-5.6 Luna1.05M
Most searched
Gemini 4 Argon2M- Kimi K3658K
- GPT-6 Astra580K
- Claude Opus 5.5405K
- GPT-6.1 Sol384K
Newest
- Kimi K3.1 Pro HighspeedOct 23, 2026
DeepSeek V4 Flash Vision ExpOct 17, 2026- Muse Spark OpenOct 15, 2026
- Gemini 4 ArgonOct 13, 2026
- GPT-6.1 SolSep 29, 2026