Pricing
Pay as you go in USD per 1 million tokens. No subscription, no minimum commitment; top up from $5.
Batch API Run it as a batch and pay up to 25% lessSee batch pricing →
OpenAI
| Model | Input | Cached input | Cache write | Output |
|---|---|---|---|---|
| gpt-oss-120b-ultra | $0.24 | — | — | $1.14 Batch API: $0.18 in / $0.855 out |
Anthropic
| Model | Input | Cached input | Cache write | Output |
|---|---|---|---|---|
| claude-fable-5 | $12.00 | — | — | $60.00 Batch API: $9.00 in / $45.00 out |
| claude-opus-5-5 | $4.80 | $0.24 | — | $24.00 Batch API: $3.60 in / $18.00 out |
| claude-sonnet-5-5 | $2.40 | — | — | $12.00 Batch API: $1.80 in / $9.00 out |
| Model | Input | Cached input | Cache write | Output |
|---|---|---|---|---|
| gemini-3.5-flash | $1.80 | — | — | $10.80 Batch API: $1.35 in / $8.10 out |
| gemini-3.7-flash | $0.90 | — | — | $4.50 Batch API: $0.675 in / $3.375 out |
Other
| Model | Input | Cached input | Cache write | Output |
|---|---|---|---|---|
| deepseek-v4-flash-0731 | $0.072 | $0.018 | — | $0.216 Batch API: $0.054 in / $0.162 out |
| deepseek-v4-flash-vision-exp | $0.528 | $0.0168 | — | $1.584 Batch API: $0.396 in / $1.188 out |
| deepseek-v4-pro-0813 | $1.56 | $0.12 | — | $3.12 Batch API: $1.17 in / $2.34 out |
| deepseek-v4.1-flash | $0.24 | $0.0072 | — | $0.72 Batch API: $0.18 in / $0.54 out |
| deepseek-v4.1-flash-1017 | $0.24 | $0.0072 | — | $0.72 Batch API: $0.18 in / $0.54 out |
| glm-5.3 | $1.08 | $0.24 | — | $4.80 Batch API: $0.81 in / $3.60 out |
| glm-5.3-flash | $0.18 | $0.036 | — | $0.60 Batch API: $0.135 in / $0.45 out |
| llama-3.3-70b-instruct | $0.12 | — | — | $0.384 Batch API: $0.09 in / $0.288 out |
| llama-4-scout-17b-16e-instruct | $0.12 | — | — | $0.36 Batch API: $0.09 in / $0.27 out |
| llama-guard-4-12b | $0.216 | — | — | $0.216 Batch API: $0.162 in / $0.162 out |
| meta-llama-3.1-70b-instruct | $0.48 | — | — | $0.48 Batch API: $0.36 in / $0.36 out |
| meta-llama-3.1-8b-instruct | $0.024 | — | — | $0.048 Batch API: $0.018 in / $0.036 out |
| mimo-v2.6-flash | $0.168 | $0.0034 | — | $0.336 Batch API: $0.126 in / $0.252 out |
| mimo-v2.6-pro | $0.516 | $0.0043 | — | $1.044 Batch API: $0.387 in / $0.783 out |
| muse-glimmer-30b | $0.36 | $0.048 | — | $1.44 Batch API: $0.27 in / $1.08 out |
| nvidia-nemotron-3-super-120b-a12b | $0.102 | — | — | $0.48 Batch API: $0.0765 in / $0.36 out |
| nvidia-nemotron-3-ultra-550b-a55b | $0.60 | $0.12 | — | $2.64 Batch API: $0.45 in / $1.98 out |
| qwen3.8-27b | $0.24 | $0.06 | — | $3.00 Batch API: $0.18 in / $2.25 out |
| qwen3.8-flash | $0.1356 | $0.0169 | — | $0.4584 Batch API: $0.1017 in / $0.3438 out |
| seed-1.8 | $0.30 | $0.06 | — | $2.40 Batch API: $0.225 in / $1.80 out |
| seed-2.0-code | $0.60 | $0.12 | — | $3.60 Batch API: $0.45 in / $2.70 out |
| seed-2.0-pro | $0.60 | $0.12 | — | $3.60 Batch API: $0.45 in / $2.70 out |
How billing works
- Token counts come from the model provider's own usage report for each request.
- Reasoning ("thinking") tokens are billed as output tokens.
- Requests that fail upstream are not charged.
- Each request is charged in fractions of a cent and deducted from your prepaid balance.