Frontier intelligence. Open weights. A fraction of the price.
NexPontis is the open-weights AI gateway: one OpenAI-compatible endpoint at https://api.nexpontis.com/v1 for the best open models — DeepSeek, Kimi, GLM, MiniMax and more — with transparent per-token pricing, wholesale volume quotes, and smart multi-source routing with automatic failover. 68 curated open models live. Prepaid, per-token billing; no subscription.
Featured frontier models
- zhipu/glm-5.3 — $1.40 input / $4.40 output per 1M tokens
- deepseek/deepseek-v4-flash — $0.140 input / $0.280 output per 1M tokens
- deepseek/deepseek-v4.1-flash — $0.150 input / $1.20 output per 1M tokens
- moonshot/kimi-k3 — $3.00 input / $15.00 output per 1M tokens
Live prices (USD per 1M tokens)
| Model | Input | Output |
|---|---|---|
| baai/bge-m3 | $0.010 | $0.000 |
| baai/bge-reranker-v2-m3 | $0.010 | $0.000 |
| baidu/ernie-4.5-vl | $0.420 | $1.25 |
| black-forest-labs/flux-2-dev | $0.000 | $0.000 |
| black-forest-labs/flux-2-flex | $0.000 | $0.000 |
| black-forest-labs/flux-2-max | $0.000 | $0.000 |
| black-forest-labs/flux-2-pro | $0.000 | $0.000 |
| bytedance/seed-1.6 | $0.300 | $2.00 |
| bytedance/seed-1.6-flash | $0.080 | $0.300 |
| deepseek/deepseek-r1-0528 | $0.500 | $2.18 |
| deepseek/deepseek-v3-0324 | $0.250 | $1.00 |
| deepseek/deepseek-v3.1 | $0.270 | $1.00 |
| deepseek/deepseek-v3.2 | $0.260 | $0.380 |
| deepseek/deepseek-v4-flash | $0.140 | $0.280 |
| deepseek/deepseek-v4-flash-0731 | $0.440 | $1.32 |
| deepseek/deepseek-v4-flash-vision-exp | $0.440 | $1.32 |
| deepseek/deepseek-v4-pro | $1.74 | $3.48 |
| deepseek/deepseek-v4-pro-0813 | $1.32 | $3.96 |
| deepseek/deepseek-v4.1-flash | $0.150 | $1.20 |
| google/flash-image-2.5 | $0.000 | $0.000 |
| google/gemma-3-12b-it | $0.060 | $0.200 |
| google/gemma-3-27b-it | $0.090 | $0.160 |
| meta/llama-4-scout | $0.180 | $0.590 |
| minimax/minimax-m1 | $0.550 | $2.20 |
| minimax/minimax-m2 | $0.300 | $1.20 |
| minimax/minimax-m2.1 | $0.300 | $1.20 |
| minimax/minimax-m2.5 | $0.300 | $1.20 |
| minimax/minimax-m3 | $0.600 | $2.40 |
| mistral/devstral-2-123b | $0.050 | $0.220 |
| mistral/ministral-3-14b | $0.200 | $0.200 |
| mistral/ministral-3-3b | $0.100 | $0.100 |
| mistral/ministral-3-8b | $0.150 | $0.150 |
| mistral/mistral-large-3-675b | $0.500 | $1.50 |
| mistral/mistral-small-4-119b | $0.150 | $0.600 |
| moonshot/kimi-k2.5 | $0.600 | $3.00 |
| moonshot/kimi-k2.6 | $0.950 | $4.00 |
| moonshot/kimi-k2.7-code | $0.950 | $4.00 |
| moonshot/kimi-k2.7-code-highspeed | $1.90 | $8.00 |
| moonshot/kimi-k3 | $3.00 | $15.00 |
| nvidia/nemotron-3-embed-1b | $0.010 | $0.000 |
| nvidia/nemotron-3-embed-8b | $0.030 | $0.000 |
| nvidia/nemotron-3-super | $0.200 | $0.800 |
| nvidia/nemotron-3-ultra | $0.800 | $2.60 |
| nvidia/nemotron-3.5-lightning | $0.100 | $0.950 |
| openai/gpt-oss-20b | $0.040 | $0.200 |
| qwen/qwen3-235b | $0.200 | $0.600 |
| qwen/qwen3-30b | $0.090 | $0.300 |
| qwen/qwen3-coder-plus | $1.40 | $5.00 |
| qwen/qwen3-max | $1.60 | $6.40 |
| qwen/qwen3-next-80b | $0.140 | $1.40 |
| qwen/qwen3-vl-235b | $0.220 | $0.880 |
| qwen/qwen3-vl-32b | $0.500 | $1.50 |
| qwen/qwen3.5-397b | $0.600 | $3.55 |
| qwen/qwen3.8-2.4t | $1.90 | $5.70 |
| qwen/qwen3.8-27b | $0.350 | $2.75 |
| tencent/hy-mt2-lite | $0.044 | $0.177 |
| tencent/hy-mt2-plus | $0.074 | $0.295 |
| tencent/hy-mt2-pro | $0.074 | $0.295 |
| tencent/hy3 | $0.132 | $0.528 |
| xiaomi/mimo-v2.5-pro | $0.435 | $0.870 |
| zhipu/glm-4.5v | $0.600 | $1.80 |
| zhipu/glm-4.6v | $0.300 | $0.900 |
| zhipu/glm-4.7 | $0.450 | $2.00 |
| zhipu/glm-5 | $0.800 | $2.55 |
| zhipu/glm-5.1 | $1.26 | $3.96 |
| zhipu/glm-5.2 | $1.40 | $4.40 |
| zhipu/glm-5.3 | $1.40 | $4.40 |
| zhipu/glm-5.3-flash | $0.150 | $0.500 |
Wholesale pricing for production volume
NexPontis is built for teams: transparent per-token list prices, and wholesale rates by quote for committed volume — below other model-API pricing. Capacity planned around your tokens/day and peak RPM. Request a volume quote — response within one business day.